Skip to main content

Full text of "ERIC ED431281: Tuning in to the 21st Century through Assistive Technology: Listen to the Music. Proceedings of the RESNA '94 Annual Conference (Nashville, TN, June 17-22, 1994). Volume 14."

See other formats


DOCUMENT RESUME 



ED 431 281 



EC 307 265 



AUTHOR 

TITLE 



INSTITUTION 

ISBN 

ISSN 

PUB DATE 
NOTE 
PUB TYPE 
EDRS PRICE 
DESCRIPTORS 



Binion, Mary, Ed. 

Tuning in to the 21st Century through Assistive Technology: 
Listen to the Music. Proceedings of the RESNA '94 Annual 
Conference (Nashville, TN, June 17-22, 1994) . Volume 14. 

RESNA: Association for the Advancement of Rehabilitation 
Technology , Arl ington , VA . 

ISBN-0- 932101-34 -8 

ISSN-0883-4741 

1994-06-22 

592p. 

Collected Works - Proceedings (021) 

MF03 /PC24 Plus Postage . 

Adults; Agency Cooperation; Appropriate Technology; 
♦Assistive Devices (for Disabled) ; *Augmentative and 
Alternative Communication; Delivery Systems; *Disabilities ; 
Dyslexia; *Educational Technology; Elementary Secondary 
Education; Employment; Federal Legislation; Gerontology; 
Individualized Education Programs; *Legal Responsibility; 
Mobility Aids; Older Adults; Public Policy; Rehabilitatijbn 
Programs; Robotics; School Responsibility; Special 
Education; Teacher Education; Transportation 



ABSTRACT 



Papers in this conference proceedings focused on the 
progress and potential of assistive and rehabilitation technology for 
individuals with disabilities and ways that RESNA members could help these 
ideas to be realized. Presentations were delivered on the following topics: 
(1) service delivery and public policy issues; (2) personal transportati<fc; 
(3) augmentative and alternative communication; (4) drooling; (5) 
quantitative assessment; (6) special education, including telephone 
technology for sensory integration, assessing predispositions to technoL^y 
use in special education, explaining legal ramifications of the appropri^t# 
application of assistive technology in the Individualized Education Progifea^, 
voicing dyslexia remediation, speech evaluation of habilitation training. 
children with hearing impairments, and innovative interagency collaborat:^dlifi; 
(7) technology transfer; (8) sensory aids; (9) wheeled mobility and seating; 
(10) electrical stimulation; (11) computer applications; (12) rural 
rehabilitation; (13) assistive robotics and mechatronics ; (14) job 

accommodation and employment issues; (15) information networking; (16) 
gerontology; (17) international appropriate technology; (18) Easter Seal 
Student Design Competition; and (19) Whitaker Student Scientific Paper 
Competition. Presentations include references . (CR) 



********************************************************************************* 

* Reproductions supplied by EDRS are the best that can be made * 

* from the original document . * 

******************************************************************************** 



best copy available 




Twdng in to the 21st Century tbmugh 




U.S. DEPARTMENT OF EDUCATION 
Office of Educationaf Research and Improvement 
EDUCATIONAL RESOURCES INFORMATION 
y CENTER (ERIC) 

Q^his document has been reproduced as 
received from the person or organization 
originating it. 

□ Minor changes have been made to 
improve reproduction quality. 



Points of view or opinions stated in this 
document do not necessarily represent 
official OERI position or policy. 



"3. S ' 

« s I 













PERMISSION TO REPRODUCE AND 
DISSEMINATE THIS MATERIAL HAS 
BEEN GRANTED BY 






er|c* 



TO THE EDUCATIONAL RESOURCES 
INFORMATION CENTER (ERIC) 









I 



PROCEEDINGS 
of the 

RESNA ’94 
Annual Conference 

Tuning in to the 
21st Century Through 
Assistive Technology 

Listen to the Music 

June 17-22, 1994 

Nashville Convention Center 
Nashville, Tennessee 

Mary Binion 
Editor 

Randall Dickman 
Jan Coatney 
Conference Co-Chairs 



RESNAPRESS 




3 



Proceedings of the RESNA '94 Annual Confraence 
Nashville, Tennessee 
June 17-22, 1994 

Volume 14 
ISSN 0883-4741 
ISBN 0-932101-34-8 

Copyright © 1994 
RESNA 

All rights reserved. This book or any part thereof nuiy not be 
reproduced in any form without the permission of the publisher. 




1700 North Moore Street, Suite 1540 
Arlington, VA 22209, (703) 524-6686 



Printed in the United States of America 
Dependable Printing Company, Inc. 



Foreword 



On behalf of the RESNA Board of Directors and this year’s RESNA Local Committee we 
welcome you to Nashville, Home of the Grand Ole Opry, for the 1994 RESNA Annual 
Conference. 

Once again, the conference promises to provide a unique opportunity for consumers, vendtws, 
service providers, and researchers from all over the world to meet together and exchange 
information about assistive technology, present and future. From the (serious) paper 
presentations to the (fun) Dance Nite, the Conference offers the attendees a number of 
opportuitities to share information, renew old fiiendships, and make new ones. 

We wish to thanlc everyone who worked to make this year’s conference a success. We 
particularly wish to recognize Susan Leone for her enthusiasm, expertise, and tirelessness in 
monitoring the hundreds of details necessary to bring the conference together. 

Have a great conference. See you again next year! 



Jan Coamey and Randall Dickman 
RESNA ’94 Conference Co-Chairs 

Gregg C. Vanderheiden 
RESNA President 



5 

O 

ERIC 



III 



I 



Preface 



Welcome to RESNA ’94 "Tuning in to the 21st Centuiy through Assistive Technology-Listen 
to the Music." As we prepare to become involved in this conference, let us remind ourselves 
about the symphony of assistive and rehabilitation technologies and the ^propriateness of 
this theme. 

We are the instruments. The assistive and rehabilitation technologies are our tools on which 
to play. They are the way that we bring success to ourselves and those we serve. We are 
the voice in Ae assistive technology song. 

We are the composers. We are writing the elements to define who we are. We are creating 
the music that awakens new ideas. With the changes that are occurring in the assistive and 
rehabilitation technology field, we are searching for the pivotal moment, when the right 
choice will bring us closer to achieving our goals. 

We are the musicians. We tailor our ideas successfully to meet the conditions of the 
changing world. Sometimes we improvise. The key to our success is that we reach out, 
extend our music to it’s boundaries, and then do it over again. We work together to extend 
ourselves and our talents. 

We are the conductors. We interpret what has been written. We take a group of divergent 
individuals and blend their talents and skills to create a symphony orchestra. We determine 
the emphasis-the crescendo. We are constantly seeking better ways, and we are not 

intimidated. 

We have many reasons to celebrate and many notes to sing. Some of us respond bett^ to 
the soft int nn ari nns , while Others prefer the rhythmic, pulsating beat Now you must chose 
to write and play your songs. I hope this concert provides you with an incubator for new 
i^pa s, an opportunity to develop action plans, and a time to build new coalitions. Enjoy! 



Mary Binion 

Chair, Scientific Program and Special Interest Groups 



6 

O 

ERIC 



V 



RESNA Board of Directors 



Executive Committee 
President 

Gregg Vanderheiden, PhD 
Trace R&D Center, University of Wisconsin 
President Elect 

Clifford E Brubaker, PhD 

University of Pittsburgh School of Health/Rehab Sciences 
Secretary 

Alexandra Enders, OTR 
RTC Rural, University of Montana 
Treasurer 
Jan C Galvin 

National Rehabilitation Hospital 
At Large Members 

Rachel Wobschall, MIM 
Minnesota Star Program 
Paul N Hale Jr, PhD PE 
Louisiana Tech University REC 
Immediate Past President 
Douglas A Hobson, PhD 

University of Pittsburgh School of Health/Rehab Sciences 
Board of Directors 

David Cooper MS Kin, Sunny Hill Hospital 
Albert M Cook, PhD PE, California State University 
David F Law Jr, Woodrow Wilson Rehab Center 
Donald R McNeal PhD, Rancho REC 
Beth Minco PhD, AI duPont Institute ASEL 
Jean Minkel MA PT, Helen Hayes Hospital 
Kathleen R Riley BSPT, Durham, North Carolina 
Carol Sargent BSl^ OTR/L, Houston, Texas 
Michael Silverman, Pin Dot Products 

Elaine Trefler, OTR Med FAOTA, University of Pittsburgh School of Health/Rehab Sciences 

Robert Van Etten MSE BSE, Adaptive Living 

Gerald Weisman, MS, Vermont REC 

Jody J Whitmyer, Whitmyer Biomechanix Inc 

Past Presidents 

Richard A Foulds PhD, AI duPtint Institute ASEL 

C Gerald Warren MPA, C Gerald Warren & Associates 

Sheldon R Simon MD, Ohio State University 

Dudley S Childress PhD, Northwestern University 

Morris Milner PhD PEng CCE, Hugh MacMillan Rehabilitation Centre 

Donald R McNeal PhD, Rancho Los Amigos REC 

Colin A McLaurin ScD, University of Virginia REC 

James B Reswick ScD, National Institute on Disability & Rehabilitation Research 



Conference Organizing Committee 



Conferaice Chairs 

Randall Dicknuui, Tennessee TAP Advisory Council 
Jan Coatnet, Mental Retardation Community Services 
Meetings Committee Chair 
Donald R McNeal PhD« Rancho REC 
Scientific Program 
Mar y Binion, Ohio Resource Ctr 
Anthony J Langton MS, South Cardina Voc Rehab 
Instructional Program 

Carol A Sargent BSEE OTR/L, Houston, Texas 
Exhibits 

Kevin Caves, Rancho REaCART 
Accessibility 

Anthony J Langton MS, South Carolina Voc Rehab 
Computer Tedi Lab 
Denis Anson OTR, Univ Washington 
Student Design Competition 
David F Law Jr, Woodrow Wilson Rehab Ctr 

Local Arrangements Committee 
Roger Blue, ARC Tennessee 
Pat Butler, Brentwood, Teimessee 
Stanley Cronk, University Tennessee-Memphis 
Errol Elshtain, Developmental Disabilities Council 
Amor Guevara, Vanderbilt University 
Vanessa Hall, Tennessee Technology Access Project 
Joel Holland, Holland Medical Equipment 
Data Howe, Coalition for Tennesseans with IMsabilities 
MoUie Ingram, Center for Independent Living 
Anastasia Koshakji, Tennessee Technology Access Project 
Michael Miller, VanderbUt University 
Cynthia Nutt-Roberts, Department of Mental Retardation 
Jackie Page, Mayor’s Advisory Committee 
Doria Panvini, Nashville, Teiuiessee 
Floyd Stewart, Center for Independent Living 
Wanda Willis, Developmental Disabilities Council 



Conference Management 
James R Geletka, Executive Director 
Susan P Leone, Meetings Director 
Terry Reamer, Conference Coordinator 




8 



VII 



RESNA Gratefully 



acknowledges conference suppwt 
from the following PATRONS 

Apple Computers, Inc 

Office of Special Education 
Audex, Inc 

Developmental Disabilities Council 
E. Power & Associates 
Everest & Jennings, Inc 
Health Purchasing Services, Inc 
Hear Our Voices 
Holland Medical Equipment 
IBM Corporation 

Martin Marietta Energy Systems, Inc 
National Easter Seal Society 
Prentke Romich Company 
Semantic Compaction Systems 
Southwestern Bell Telephone Company 
State of Tennessee Department of Human Services 
Division of Rehabilitation Services 
Whitaker Foundation 



•^recognizes the assistance of 
the following organizations 

ARC Tennessee 

Center for Independent Living of Middle Tennessee 
Coalition for Tennesseans with Disabilities 
Developmerucd Disabilities Council 
Mayor’s Advisory Committee 

Ohio Resource Center for Low Incidence & Severely Handicapped 
Rancho Los Amigos Medical Center 
South Carolina Vocational Rehabilitation Department 
Stallworth Rehab Hospital at Vanderbilt 
State of Tennessee Department of 
Mental Health & Mental Retardation 
Tennessee Commission on Aging 
Tennessee Technology Access Project 
University ofTennessee-Memphis 
University of Washington 
Vanderbilt University 
Woodrow Wilson Rehabilitation Center 



O 

ERIC 



9 



VIII 



TABLE OF CONTENTS 



SIG-01: Service Delivery & Public Policy Issues 

Social W(Hk and the Intodisciplinary Nature of Assistive Technology 2 

Tech Points: Enhancing the use of Rehabilitation Technology in Vocational Rehabilitation Agencies 5 
Engineering Snvice Provision for Vocational Rehabilitation in Tennessee 8 

Knowledge of Assistive Technology in Spinal Cord-Injured Individuals in Kentucldana 10 



Project IMPACT: Development of a Statewide Service Deliv^ Model 
through Local C^iacity Building 13 

Rehabilitation Technology Service Delivery Models in Vocational Rehabilitation 
Agencies: A Multi-Level Approach 

Tetra Society of North America: The First Three Years 19 

PROJECT IMPACT: Innovative Methods for ^viding Access to Technology 21 

Rehabilitation Engineers in Vocational Rehabilitation Agencies: A Profile 24 

Assistive Technology Usage Outcome: A Preliminary Report 27 

Modifying Commocially-Available Walkers to Meet Individual Needs 28 

Integraticm of Technology into a New Occupational Therapy Curriculum 31 

HEART - A Study on Assistive Technology in Europe 34 

Considering Quality in a Cross-Discipline Rehabilitation Engineering Service 37 

Rehabilitation Engineeiing Training - The Practical Side ^ 

In-Service and Consumer Training in Assistive Technology: A Work in Progress 43 

Home Systems Technology for Elderly and Disabled People - Present Status, 

R&D Methodology, and Future Directions ^3 

Accessibility Standtuds for Children’s Environments 48 



SIG-02: Personal Transportation 

Development of Instrumentation and Protocols to Measure the Dynamic 
Environment of a Modified Van 

Accelerations Experienced by Wheelchair Users with SQ in a Moving Van 

Rollow and Directional Stability of Vehicles Modified for the Physically Challenged 

Crash Re^nse of Wheelchair-Occupant Systems in Transport 

Wheelchair Tiedown Int^laboratory Test 

Guidelines for Wheelchair Securement and Personal Restraint 

Car Ad^tations for People with Special Needs in Europe - A Simulator Study Evaluation 
A Small Scale Driver Evaluation Vehicle 
A Human Powered Vehicle for the Physically Disabled 
Wheelchair Lift Contraindications 

Positioning and Securement of Riders and Their Mobility Aids in Transit Vehicles 



O 

ERIC 



SIG-03: Augmentative and Alternative Communication 

A Communication System Established fw a C-5 Quadriplegic and His Hearing Impaired Parents 
Development of an Assistive Technology Support Network 
A Head Gesture Recognition System for Ctanputer Access 
The Design and Development of a Computer-Based System for Assessing 
and Training Two-Dimensional Language Representation 
The Speech Translator - A Communication Aid for Severely Impaired Speakers 
TalksBac: A Predictive Adaptive Conversation System for Non-Fluent Dysphasics 
Extending the Functional Day for Powoed Wheelchair Based Technologies 
A Morphotogically-Based Word Predictor Ccx’ Swedish 

Development, Programming and Concealment of Private Messages in AAC Systems 
Design Strategies for AAC Software, Applied to the Blissprocessor 
• Do These Strategies Comply with the Users Needs? 

Translation Problems of Tose Messages in Voice Output Communication Aids 
Quantitative Indicators of Cognitive Load During Use of a Word Prediction System 
Predictive Letter Scanner for Augmentative Communication 
Communication Interaction Involving a Young Augmentative Communication 
Device Uso' and Her Partners 

Developing AAC Systems that Model Intelligent Partner Intoactions: 

Methodological Considerations 



IX 




52 

55 

58 

61 

64 

67 

70 

73 

76 

79 

82 



86 

89 

92 

95 

98 

100 

103 

106 

109 

112 

115 

118 

121 

124 

126 



^ V. 



AAC-User Theiq>ist Interactions: Pieliminary Ling uistic Observations 
and Implications for CX>MPANSION 
Changes in Interaction Among AAC Users and their High School 
Communicative Efficacy of Computerized VIC and Natural t .nnginig(» 
Improved User Interface for Wtsd Prediction 

SIG*04: Drooling 

Swallow Reminder, Saliva Pump and Ciqj with Rxed Ddivery Volume 



129 

132 

135 

138 



142 



SIG-05: Quantitative Assessment 

Objwtive Evaluation Scales for the Quantitative Assessment of Walking Function in Stroke P&tients 
Preliminary Design of an Optimized Jogging Prosthesis 

Gait Simulation and the De^ of a Paediatric Above-Knee Endn steie t ai Running Prosthesis 
Upper ExtremiQr Net Joint Fences and Moments During Wheelchair Propulsion 
Differential Preraure Walking Assist 

Measuring Leg Motion Changes Fdlowing Vertical Vestibular Stimulation: A Case Study 
FUnctimial Assessment of a Tricqrs Orthosis fin CS/C6 Tetraplegia 
The Analytical Inadequacies of TTeadmill-Mounted Force Platfinms 
Accuracy of Upper-Extremity Motion by Healthy and Disabled Subjects 
An Instrument for Testing Aimed Movements and Applied Forces 
The Comparative Evaluation of Isotonic and Isokinetic Modes of Testing with 
Ergonomic and Rehabilitation Poiqrective 
A Validation Study of a Lift Simulator During igninfatini Lifts 
The Effects of Lo^ Mode, Speed of Lift on the Power Goieration, Absorption 
and Transfer During a Multi-Link Coordinated T ifring Task 



146 

149 

152 

155 

158 

161 

164 

167 

170 

173 

176 

179 

182 



SIG-06: Special Education 

Education and Assistive Technology: Florida’s Model 

ACCESS: Special Educators on Line 

Telqrhone Technology fin' Sensory Integ ration 

Assessing Prediqrositions to Technology Use in Special Education: 

Music Education Majors Score with the "Survey on Technology Use" 

ExpliHing Legal Ramifications of the Appropriate J^lication of Assistive 
Technology in the Individualized Education Program 
Studoits Who Use Assistive Technology: Science Gets Ahead of the Law 
Voicing Dyslexia Remediation 

Speech Evaluation of Habilitation Trainmg of Heariqg-Impaired Children in Shanghai, China 
Innovative Interagency Collaboration: Promoting the inciiKinn of a Student 
Using Augmentative Communication in Regular 



186 

188 

191 

194 

197 

200 

203 

206 

208 



SIG-07: Technology Transfer 

Applying Assistive Techndogy Concepts to Non-Disability Problems 
Study of Accessible Microwave Oven Design 
Accessibility Evaluation of Current Television Design Trends 
Siqipon for Technology Transfer in the Ontario Rehabilitation Technology rnnsm tiiim 
Assistive Technology Training: An Interdisciplinary Approach 
RERC on Technology Evaluation and Transfer Program Access and Value Added 
The VA Rehab R&D Technology Transfw Process 
Produa Design for People with Disabilities: A New Graduate Course at 
the University of Teiuiessee-Memphis 
RR&D’s Bag of Technology Ttansfer Tools 
Design Issues for People with Special Needs - A Study with Drivers 
User Involvement in Assessment and User Inf luenee in Standardization of 
Consumer Products and Assistive Technology 



212 

214 

217 

220 

223 

225 

228 

231 

234 

237 

240 




X 



11 



SIG-08: Sensory Aids 

Eyewear that Precisely Controls light Intensity Levels at the Eye 

A L ig nid Crystal Adjustable Power Lens Magnifier for Perscms with Low Vision 

An Instruction Sequence for Elderly Posons witfi Macular Degeneration Using Low Vision Devices 

Evaluation of a Robotic Fingerspelling Hand 

Survey of Deaf-Blind Technology Needs 

Improved PCTcepdon of Music, by Vibro-Tactile Means, for Deaf People 
FuncticMial Assessment of Vision Undm- Poor Lighting and Contrast: The Skill Card 



244 

247 

250 

252 

255 

258 

261 



SIG-09: Wheeled Mobility and Seating 

Uniaxial and Hydrostatic Loading at the Core of a Gel Buttock Model 
Blood Lactate Re^nse of Wheelchair Racers to Maximal Exmcise 
Biomechanical Aqiects on the IRV Mobile Headrest 
Surface Assessment Devices for Accessilnlity 
Assessment of Outdow Environments for Accessibility 
A G-Tube Pump Mount for Powct and Manual Mobility Aids 
Modifying the Subasis Bar to Enhance its Perfomance 
Case Studies of Performance Using Integrated and Distributed Controls 
A Fbur Degree of Freedom Passive Digitizer for Contoured Seating Surface 
Technology Identification fe a Non-Invasive Spinal/Pelvic Alignment Monitoring 
System for Individuals Seated in Prasonal Wheeled Mobility Devices 
A Low Cost, Adjustable Vacuum Form Unit 
Customized Integrated Control System 
The Evolution of a Modular Augmentative Mobility System 
A Tracing Head/Bairier Avoidance Scheme Cw a 3-Axis Automatic Shape 
Sensing Machine for Special Seating 
Trail Guides with Unive^ Access Information 
A Flexible Integrated Access System 

Safety Versus Functionality, A European Discussion on Standards for 
Wheelchairs and Related Products 
Effect of Camber on Wheelchair Stability 
Determination of Static Stability of Bowel Care-Showo' Chairs 
The Effect of Reaching and Leaning on Wheelchair Stal^ty 
Manual Wheelchair ISO-ANSl/RESNA Fatigue Testing Experience 
The UVA/MR Maintenance-Free Wheelchair Tire 
Portable Head Support 

Three Custom AdEqrtations for a Child with Segmental Dystonia 
A Wheelchair-Mounted Infant Care Seat 

A Powered Mobility Aid for Pre-School Children with Cerebral Palsy 
Nfobility and Mobility Training fe Severely Disabled Children: Results of 
the "Smart” Wheelchair Project 
Wheelchair Battery Overdischarge & Overcharge Problems 
New Technology for Wheelchair Batteries 

An Automatically-Guided Powered Wheelchair for the Severely DisaUed 
Tntfiiigftnt Onuiidirectional Wheelchair with a Flexible Configurable Functionality 
Initial Evaluatimi of Power Add-On Units for Lightweight Wheelchairs 
Electromagnetic Compatibility (EMC) of Powoed Wheelchairs and Scooters 



266 

269 

272 

275 

278 

281 

284 

287 

290 

293 

296 

298 

301 

304 

306 

309 

312 

315 

318 

321 

324 

327 

330 

333 

336 

339 

341 

344 

347 

350 

353 

356 

359 




SIG-10: Electrical Stimulation 

Transcutaneous Electrode Impedance Measurement System 

An Instrumented Grasp Sensor for Quantitative Evaluations of Neuroprosthetic Hand Grasp Systems 
Strength Acquisition Techniques in Paralyzed Quadric^ Muscle using 
Functional Electrical Stimulation 

Volitional Exercise Plus Electrical Stimulation: Significantly Improved Muscle Perfomance 
The Rehabilitation of Gait Aftm- Stroke: A Comparison between Conventional 
Ther^y and Muldchaiuiel Functional Electrical Stimulation Therapy 
Paraplegic Walking with Functional Neural Stimulation 

XI j 2 



362 

364 

367 

370 

373 

376 



SIG-11: Computer Applications 
"BEST" Products for Computer Access 

Scanning Assessment Tool Assessing Selection Control Techniques 
Shifting Priorities in Elevelopment of an ABLEDATA CD-ROM 
Access to Business Gr^hics for Blind People 
Scanning the Windows Desktop without Mouse Rmnlatinn 

A Comparison of Conputer Access Devices for Persons with High Level Spinal Cord Injuries 
Alternative Access to Assistive Devices 

An Integrated Remote Voice-Accessed Point of Care Solution for Pharmacy Onlers 
Computer Use in Training and Education - A Meta Analysis of S ingle Subject Design 
Modified Super Nintendo Controller 
Mouse Breakout: Relocation of Mouse Switch Functions 
Function^ Evaluation of the Datahand® Key Entry System Us» Experience 
Evaluation by Questionnaire 

Towards Gesture Recognition for the Physically Impaired 

Hand Crmfiguration Pre-Processing Tool for Sing l .iingiing<» Recognition 

DEFIE: An Intelligent and Flexible Integrated Environment fw Disabled and Elderly Petrie 

Cognitive Orthotic Shell 

To Evaluate or Not to Evaluate? - That is the Question 

The Mercator Project - Providing Access to X Windows for Computer Uscts Who are Blind 
SIG-12: Rural RehabiUtation 

Developing a Research Agenda to Initiate Rehabilitation Services fw 
Migrant and Seasonal Farmworkers 
The Grippen A One-Handed Cattle Castrator 
An Affordable Lift System for Farmos with Mobility Impairments 
International Agricultural Rehabilitation 

SIG-13: Assistive Robotics and Mechatronics 

Use of a Rehabilitation Robot in a Classroom Setting 

Impact of Mechalronic Systems as Vocational Enables 

The Design of an Integrated Interface to an Educational Robotic System 

Evaluation of the RAID Workstation 

Eye Movement Control of the Manus Manipulator 

EMG Patton Analysis fw Prosthesis Arm Qmtrol 

Vibrotactile Feedback for Dextrous Teleoperation 

Feeding the Physically Challenged Using Coopoative Robots 

A Robust, Self-Diagnosing Sensing Methodology fw ^rplication to Smart Wheelchairs 

The Assistive Research and Technology Wheelchair Mounted Robotic Arm: Prototype Review 

The Use of the Soft Arm for Rehabilitation and Prosthetics 

Motion Simulation for an Improved Orthosis 

SIG-14: Job Accommodation 

A Graphical User Intoface Incorporating Braille and Musical Sounds to 
Accommodate Blind Communications Assistants 
An Analysis of Woildng Postures of Manual Wheelchair Users in the Office Environment 
Business and RehabiUtation: A Partnership on Worksite Accommodatirxi 
Modification of Industrial Hose Nozzle and Work Are for Worko- with Arthritis 
Long-Armed Bottom Wip» with a Detachable Handle 
The Successful Integration of Assistive Technology and Job Development 
for an Individual with a DisabiUty: A Case Study 
Assessment and Training Devices for Work Placements at the Hissom Memorial Ce nter 



380 

383 

386 

388 

391 

394 

397 

400 

403 

406 

409 

411 

414 

417 

420 

423 

426 

429 



432 

433 
436 
439 



442 

445 

448 

451 

454 

457 

460 

463 

466 

469 

472 

475 



480 

483 

486 

487 
490 

492 

495 




1 3 



XII 



SIG'15: Information Networking 

Satim n - A foit i pin»ri7i».ri Information and Refenal System for Assistive Teclmology 
and Individuals with Disabilities 

Analysis of ABLEDATA Information Requests: Implications for Assistive Technology 
Infcmnation Dissemination 

Interdisciplinary, Multilevel Training in Assistive Techndogy 
Use of Multiple Parallel Interface Strategies to Create a Seamless Accessible 
Interface for Next-Generation Information systems 

SIG-16: Gerontology 

Use of Technotogy to Improve Service to Older Adults: A Computerized 
Community Caie Coodination Network 

Difficulties and Dqiendence in Bathing: Interviews with Bathers and Care-Providers 
Applicfl b‘ii*y of Grab Bar Accessibility Standards to Meet the Needs of Older Peqile 
Oppwtunities fw High Technology R&D to Fulfill Needs of the Eldaly 

SIG'17: International Appropriate Technology 

A Low Cost Vacuum Consolidation System to Facilitate Seating Evaluations 
Managed r o «" r<*ritinn in Wheelchair Production: A Transition from Plarmed to Market Economy 
Low Cost Pressure Measurement Technique for Individuals with Spinal Cord Injury 
Advancement of Appropriate Rehabilitation Technology in Indonesia 



500 

502 

505 

508 



512 

515 

518 

521 



526 

529 

532 

535 



ggctpr Seal Student Design Competition 

Development of an Efficient Computer Interface for Persons with Mobility Impairments 

Intoactive Scimce Experiments Allowing the Inclusion of Students with Disabilities 

Multisensory Interface to Allow Blind User Access to Graphics 

The Design of a Foot Actuated Control Intoface 

Switch Activated Dispenser for Packaging 

The Design of A Hydraulically Powered Elevating Toilet Seat 

A New Switch Actuator Design that Enhances Proiwioceptive Feedback in The User of Limb Position 



543 

546 

549 

552 

555 

558 

561 



Whitaker Student Scientific Paper Competition 

The Fffry* of a Word Prediction Feature on Text Generation Rate 

PUot Study of the Effects of Visual Feedback on the Graq;)ing Function of Prosthetic Hands 
F r pprii"^r»tfl’ Determination of Stress Distribution Due to Wheelchair Dynamic 
Loads for Use with Finite Elements 

The Applimbili*y of Neural Networks in an Autonomous Mode Selection System 
Tvcign Criteria for Obstacle Avoidance in a Shared-Control System 



569 

572 

575 

578 

581 



Author Index 



585 




14 

XllI 



SIG-01 

Service Delivery and Public Policy 



SOCIAL WORK AND THE INTERDISCIPLINARY NATURE OF ASSISTIVE TECHNOLOGY 



M. Diane Calloway, Ph.D, Social Work, Utah State University 
Craig C. Shaffer, Social Work Intern Student 
Assistive Technology Program Utah State University 



ABSTRACT 

Assistive technology has experienced rapid changes 
in philosophy and practice, from instituticmalization 
to full inclusion and from isolation to independence. 
As a result professionals from both technology and 
htunan services are a part of an interdisciplinary 
team facilitating this independence. Where is 
Social Work’s place in the Assistive Technology 
interdisciplinary team? It can be an en^owering 
force, one which links the consumer and the in- 
terdisciplinary team with the resources to better 
achieve personal independence of the people with 
disabilities. They can do this with a knowledge 
base unique to the social work profession. The 
effects of the social environment on the consumer’s 
ability and willingness to use their assistive technol- 
ogy can be addressed by the interdisciplinary team 
more effectively with the addition of the social 
work profession. 

BACKGROUND 

The last decade has witnessed monumental changes 
not only in the types of services provided to people 
with disabilities but the basic {Mlosophy which re- 
flects the values that we place on people with or 
without disabilities. Rather than tolerate diversity 
we are celebrating it with a philosophy encom- 
passed in the terms ’’normalization" and "inclu- 
sion." The philosophical changes that have oc- 
curred in disabilities services and practices are well 
reflected in the language and provisions of recent 
legislation. Past legislation and the more recent 
Americans with Disabilities Act have all attributed 
to consumer responsiveness, consumer directedness, 
consumer participation and decision- making, and 
the right to access public services. As a result, 
technology is utilized to assist people in becoming 
independent. The availability of sophisticated infor- 
mation, down from the professional to the end user, 
coupled with increased concern for individuality, 
independence, and participation has created a 
paradigm shift in the disability field. Perhaps the 
most important change has come in the role of the 
consumer from a passive, grateful recipient to an 




active participant in determining need, making 
intervention decisions, and commitment. 

Assistive technology is a vital interest to the family, 
both nuclear family and extended, and society in 
general. The role of technology and disabilities 
affect our service institutions, hospitals, schools, 
rehabilitation agencies, and job placement as well. 
These are issues that the rehabilitation engineer, the 
supplier, and the clinical professions have not had to 
deal with in the past and have had little tr aining or 
experience in. 

Assistive technology services is truly interdisciplin- 
ary and rather than expanding each clinical or 
technical discipline to include expertise in the social, 
environmental, and societal implications of technol- 
ogy we should reach out to those professions that 
have traditionally addressed these issues, such as 
social work, social psychology, family development, 
and community organization. 

It is with this premise that the Assistive Technology 
services at Utah State University was prompted to 
establish an interdisciplinary careers program. 
Included in this program the first year were the 
technological professions of assistive technology, 
engineering, and computer science; and the h uman 
service professions of psychology, communicative 
disorders, special education, and occupational thera- 
py. The profession of social work was added the 
following year in an attempt to expand the tpams 
effectiveness. 

STATEMENT OF PROBLEM 

As we have mentioned earlier, the utilization of 
advanced technology can help make the lives of 
people with disabilities considerably more indepen- 
dent. An interdisciplinary team is the most effective 
model to facilitate this independence. There has 
been a great increase in the information that is 
available to both the professional and the consumer. 
The consumer is now an active participant in this 
decision-making process. An interdisciplinary team 
works together with the consumer and shares infor- 
mation to better provide for the consumer’s needs. 



2 



RESNA’94 • June 17-22, 1994 



SOCIAL WORK ROLES 



In the Assistive Technology at Utah State Universi- 
ty we have found that just using technical people 
from different disciplines is falling short of rea- 
ching the goal of a fully functioning interdisci- 
plinary team. In order to facilitate the indepen- 
dence of consumers, we must be more aware of the 
effects of their social environment on their ability 
and willingness to use needed technology. Adding 
team members from the social sciences, such as so- 
cial work, facilitates a move toward understanding 
better the consumer in their environment. 

The problem to date has been that the bigger focus 
has been on the technical independence for the 
consumer. We have yet to fully understand the 
reciprocal transactions that take place between the 
consumer and the many systems in his/her environ- 
ment that make independence complete. To achieve 
our aspirations in this regard we need to find a way 
to incorporate this needed sensitivity into the 
interdisciplinary team. 

APPROACH 

Our premise is that the social work profession con- 
tributes to the interdisciplinary nature of assistive 
technology because it is a multi-skilled profession. 
Social workers are knowledgeable in three levels of 
intervention including micro (individual), mezzo 
(families, groups, and agencies), and macro (com- 
munities and social policy). The social work 
profession is characterized by their method and 
specific knowledge base in human behavior and the 
social environment. A social systems model, in 
part, is where oiu* knowledge and skills are derived. 
This knowledge building is focused on the interplay 
between the person and his/ her environment. The 
concepts and techniques of an empowerment ap- 
proach can promote a person interacting more ef- 
fectively in his/her environment. 

The empowerment approach is a process for in- 
creasing personal, interpersonal, or political power 
so that consumers can take action to improve their 
life situations. It allows the consumer to develop a 
sense of personal power, and ability to affect 
others, and an ability to work with others to change 
social systems. Empowerment techniques used to 
increase independence include: (1) accepting the 
consumer’s definition of the problem, (2) identify- 
ing and building upon existing strengths, (3) teach- 
ing specific skills, (4) mobilizing resources in the 



consumer’s environment, and (5) advocacy support 
groups. 

DISCUSSION 

The interdisciplinary team functions such that al- 
though each member may have a different set of 
skills to contribute, there is an attempt to thoroughly 
share the information during assessment and post 
planning, and that a consensus is obtained in order 
for the team to function as a whole. Further, there 
is generally a case manager who is qualified to 
summarize, extricate and convey findings to the cli- 
ent/family/agencies and provide recommendations 
for therapies or for providing solutions to specific 
problems. It is in this role of either case manager 
or consultant that the social worker could play a sa- 
lient role. A social worker, using an empowerment 
approach, can make a unique contribution in the 
following three ways: (1) focusing on funding 

issues which entail facilitating families in making 
use of community resources, including services and 
financial assistance; (2) a liaison role keeping lines 
of co mmuni cation open between agencies, family, 
and client; and (3) the worker can also work with 
agencies, local communities, and legislators to make 
technology more available to clients. In addition, 
social workers along with other professionals, can 
provide counseling to help clients adjust to their 
disability and to rehabilitation programs; counsel 
with the family to help them adjust; take social 
histories about the client’s family background and 
present status; and do discharge planning. 

The monumental changes in the types of services 
provided people with disabilities and the basic phi- 
losophy of these services along with the changes in 
policy have created a great need to better understand 
the client in their environment. Just as the technolo- 
gy is becoming more sophisticated so are the recip- 
rocal effects of the consumer and the environment. 
The interdisciplinary team approach has helped con- 
siderably with these issues. As we include the 
social work profession as a part of the interdis- 
ciplinary team it will function even more effectively . 

REFERENCES 

Chess, Wayne A. & Norlin, Julia M. (1988). 
Human Behavior and the social environment: A so- 
cial systems model. Boston: Allyn and Bacon. 



copy AVAIIABUE 



17 



RESNA ’94 • June 17-22, 1994 



3 



SOCIAL WORK ROLES 



Chmch, Gregory & Glennen, Sharon. (1992). The 
handbook of assistive technology. California: 
Singular Publishing Group. 

Germain, Carel Bailey. (1991). Human behavior in 
the social environment: An ecological view. New 
York: Columbia University Press. 

Gutierrez, Lorraine M. (March, 1990). Working 
with women of color: An enqrowerment perspec- 
tive. Social Work, 35(2), 149-153 

Zastrow, Charles. (1993). Introduction to social 
work and social welfare. California: Brooks/Cole 
Publishing Company. 

ACKNOWLEDGEMENTS 

Utah State University Assistive Technology Pro- 
gram 

Marvin Fifreld, Director of Center of Persons with 
Disabilities and Linda Chisholm, Coordinator Inter- 
disciplinary Training and Assistive Technology 
Program 

M. Diane Calloway, PhD 
Assistant Professor 
Department of Social Work 
Utah State University 
Logan, Utah 84322-0730 
(801) 750-2389 
FAX (801) 750-1240 




4 



RESNA’94 • 



TECH POINTS: 

Enhancing the Use of Rehabilitation Technology 
in Vocational Rehabilitation Agencies 

Anthony J. Langton, Judy L. Hughes 
Center for Rehabilitation Technology Services 
South Carolina Vocational Rehabilitation Department 



ABSTRACT 

TECH POINTS is a rehabilitation technology man- 
agement and training strategy developed for use in 
vocational rehabilitation agencies funded by the 
United States government. Under development for 
field trial, TECH POINTS is designed to provide 
rehabilitation counselors with an easy to follow 
desk reference suggesting when and how they 
should consider use of rehabilitation technology 
with their clients. 

BACKGROUND 

The Center for Rehabilitation Technology Services 
(CRTS) received a five year research grant from the 
National Institute on Disability and Rehabilitation 
Research. The focus of this grant is on rehabilita- 
tion technology applications in Vocational Reha- 
bilitation (VR) Agencies. The TECH POINTS model 
will be used as the basis for an intervention study 
assessi ng the efficacy of TECH PO INTS as a strategy 
to integrate technology into the VR process. 

Rehabilitation technology efforts in vocational re- 
habilitation (VR) agencies have concentrated on 
implementing technology service delivery capa- 
bilities within the agency or through use of outside 
service providers. These efforts have helped make 
rehabilitation technology services more available, 
but they haven't always shown how the use of 
technology resources and services could, or should, 
be integrated into the rehabilitation process. Even 
with the mandate to include rehabil itation technol- 
ogy services as part of the vocational rehabilitation 
process since 1986, these services are still not 
effectively integrated into regular case manage- 
ment activities (Langton, 1 991 ). 

Access to rehabilitation technology services is in- 
fluenced significantly by the knowledge and capa- 
bility of vocational rehabilitation counselors to use 
these resources and services with individuals on 
their caseloads (Phillips, 1 992; Rice, 1 990). 
Improving the integration of rehabilitation technol- 
ogy into the VR case service process was the 
objective of project activities initiated by the Cen- 
ter of Rehabilitation Technology Services. A con- 



ceptual model, TECH POINTS, was developed to 
determine when and where technology resources 
should be utilized in the VR process tion (Langton, 
1991). This model identified nine specific points 
within the rehabilitation process where the use of 
technology should be considered. 

STATEMENT OF THE PROBLEM 

The original TECH POINTS model was well re- 
ceived as an effective illustration of how rehabili- 
tation technology services could be integrated into 
the VR process. Further development revealed 
some concern with the expectations of the rehabi li- 
tation counselor and the compatibility that the 
system would have with different VR agency struc- 
tures. It was determined that in order for TECH 
POINTS to be an effective strategy, the system had 
to be designed for use within any state-federal 
vocational rehabilitation program, with only a mi ni- 
mal amount of paperwork or procedural changes. 

Preliminary work with the model suggested that 
modifications were needed to: 

1 . Better integrate the TECH POINTS into the VR 
case service process, 

2. minimize the number of TECH POINTS, 

3. make modifications to conform to the 1 992 
Rehabilitation Act Amendments. 

APPROACH 

Working with a study group of rehabilitation case 
management staff and technology specialists, the 
original model was modified to better conform to 
the VR case service process. The original nine 
TECH POINTS were reduced to seven, with adjust- 
ments made for changes created by the 1992 
Rehabilitation Act Amendments. Still employing a 
series of "critical junctures" in the rehabilitation 
process, TECH POINTS now provides the voca- 
tional rehabilitation staff with easy to follow guide- 
lines for systematic consideration of rehabilitation 
technology services for VR clients. 



June 17-22, 1994 



RESNA ’94 



5 



TECH POINTS: Enhancing the Use of Rehabilitation Technology in Vocational Rehabilitation Agencies 



Figure One illustrates the seven TECH POINTS 
within the overall vocational rehabilitation pro- 
cess. 




Since TECH POINTS is based on the existing reha- 
bilitation process which is generally consistent 
throughout all VR agencies, it should integrate 
rehabilitation technology services into regular case 
service activity. The revised TECH POINTS model 
is a continuous process which begins with consid- 
eration of rehabilitation technology needs that a 
prospective client might have and systematically 
follow that individual throughout their rehabilita- 
tion process. 

TECH POINTS 

The following information briefly describes the 
seven TECH POINTS. Reference is given to the case 
status codes numbers, such as (02), which are in 
common use by VR agencies. 



TECH POINT Referral/Application 

( 00 , 02 ) 

Initial technology-needs screening of all applicants 
for VR services, which includes completion of a 
rehabilitation technology needs profile for those 
appi icants who use technology or where a technol - 
ogy-related need has been identified. 



B|ST COPY AVAILABLE 
ERIC 6 




TECH POINT Extended Evaluation 

( 06 ) 

Suggestions given for including rehabilitation tech- 
nology services in comprehensive assessment and 
vocational evaluation services, with use of a rehabili- 
tation technology assessment recommended for in- 
dividuals with identified technology-related needs. 



TECH POINT Plan Development 

( 10 , 1 2 ) 

Suggested approaches on how to i ncl ude appropri - 
ate use of rehabilitation technology to achieve 
vocational goals as part of the individually written 
rehabilitation program (IWRP). 

TECH POINT Services 

( 14 , 16 , 18 ) 

Tips for including appropriate use of rehabilitation 
technology in all services, including counseling 
and guidance, physical restoration and vocational 
and work adjustment training. 

TECH POINT Placement/Follow-Up 

( 20 , 22 ) 

Addresses potential need for job accommodations 
to assist in performing "essential functions" of the 
job. Includes ways to use rehabilitation technology 
with job development and in identifying feasibility 
of accommodations. 

TECH POINT Closure Points 

( 08 , 26 , 28 , 30 , 34 ) 

Review of successful and unsuccessful VR closures 
to ensure that appropriate consideration was given 
to the use of technology resources and services. 

TECH POINT Post-Employment 

( 32 ) 

Identification of rehabilitation technology inter- 
ventions needed to maintain employment. 

Technology Interventions 

Without a background in rehabilitation technology 
services, most VR counselors have difficulty deter- 
mining what rehabilitation technology services an 
individual might need. Within the TECH POINTS 
structure, descriptions are provided of possible 
"technology interventions" which could be appro- 
priate to consider. 



RESNA ’94 



June 17-22, 1994 





TECH POINTS: Enhancing the Use of Rehabilitation Technology in Vocational Rehabilitation Agencies 



Langton, A.j. & Hughes, j.K. (1 992) Back to work. 
Team Rehab , (pp. 1 4-1 8). May 1 992. 



Technology interventions include: ■ technology 
consultation, ■ vocational evaluation accommo- 
dations, ■ rehabilitation technology assessment, ■ 
home/adapted living accommodation, ■ school/ 
training accommodation, ■ job/work site accom- 
modation and ■ job development/feasibility stud- 
ies. 

DISCUSSION 

The revised TECH POINTS struaure is expeaed to 
provide the vocational rehabilitation counselor 
with an easy to follow guide for considering the use 
of rehabilitation technology with all clients. Inte- 
grating the consideration of rehabilitation technol- 
ogy into day-to-day case service practice must 
occur if these services are to be effectively utilized. 

TECH POINT should hel p to de-mystify rehabilita- 
tion technology and enable counselors to work 
more closely with technology service delivery spe- 
cialists. The VR counselor is the key for consumers 
to be able to access rehabilitation technology ser- 
vices. Effeaive use of rehabilitation technology 
depends on the counselor and consumer working 
together to recognize the potential benefit from 
these services. The importanceof initial contaaat 
point of referral and the counselor's continuing 
role i n coord inating al I rehabi litation services makes 
it essential that counselor's have a basic awareness 
of rehabi I i tation tech nology and have d i rect access 
to rehabilitation technology resources and ser- 
vices. 

Implementation of the field testing of the TECH 
POINTS model in seleaed VR agencies will begin 
in mid-1 994 and continue for approximately two 
and one-half years. Results of this study will be 
made available through CRTS. 

REFERENCES 

Rehabilitation Services Administration (1 993). The 
rehabilitation act of 1973 as amended by the 
rehabilitation aa amendments of 1 992. Washing- 
ton DC: US Dept, of Education. 

Langton, A.J. (1991) Utilizing technology in the 
vocational rehabilitation process. Proceedings of 
the 14th Annual RESNA Conference, Washington, 
DC: RESNA Press, (pp. 76-78). 

Langton, A.J., Coker, C.C. and Smith, C.A. (1 989) A 
descr i ptivestudyofrehabilitationtechnol ogy u ti I i - 
zation in state vocational rehabilitation agencies. 
lournal of Rehabilitation Administration , (pp. 45- 
50) May 1989. 



Mandeville, K.A. & Brabham, R. (1 987) The state- 
federal vocational rehabilitation program. In Parker, 
R.M. & Szymanski, E.M. (pH.) Rehabilitation Coun- 
seling: Basics and Bevond . Austin, TX:Pro-ed. (pp. 
43-71). 

Phillips, B. (1 992) Perspectives on Assistive Tech- 
nology Services in Vocational R ehabilitation: Cli- 
pnK and Counselors . (Report from Consumer Satis- 
faaion with Assistive Technology Project), Wash- 
ington DC: National Rehabilitation Hospital. 

Rice, D.(ed.) (1990) The Provision of Assistive 
Tech nology Services in Rehabilitation, Seventeenth 
Institute on Rehabilitation Issues, Research and 
Training Center in Vocational Rehabilitation, Uni- 
versity of Arkansas. 

Scherer, M. (1 993) Living i n the s tate of stuck: how 
technology impacts the lives of p ersons with dis^ 
abilities . Cambridge, MA: Brookline Books. 



ACKNOWLEDGEMENTS 

The Center for Rehabilitation Technology Services 
(CRTS) is part of the South Carolina Vocational 
Rehabilitation Department. Support for this work 
has been provided through the National Institute 
on Disability and Rehabilitation Research (NIDRR), 
U.S. Department of Education, Washington D.C. 
as part of the rehabilitation engineering research 
center grant #H1 33E20002-93. 

Assistance on the TECH POINTS research has been 
provided by Cassandra Townsend, Frank Puckett, 
Leonard Anderson, Dawn Mangum, Marcia Scherer, 
Belinda Hudson, William Will, Greg McCrew, 
Pete Howell, Kerry Mandeville, and Richard Koufal. 

Anthony). Langton 

Center for Rehabilitation Technology Services 
South Carolina Vocational Rehabilitation 
Department 
1410-C Boston Avenue 
W. Columbia, SC 291 71 



21 






RESNA ’94 



June 17-22, 1994 



7 



ENGINEERING SERVICE PROVISION FOR VOCATIONAL REHABILITATION IN TENNESSEE 



Stanley R. Cronk 

Rehabilitation Engineering Program 
University of Tennessee, Memphis 
Memphis, TN 



Abstract 

The University of Tennessee, Memphis, Rehabil- 
itation Engineering Program (UTREP) has been 
working together with the Tennessee Division of 
Rehabilitation Services (DRS) to provide job 
accommodation evaluation and support services for 
vocational rehabilitation clients throughout the state. 
Through a pre-paid contract renewed each year with 
the agency, the program has been able to establish 
two satellite sites to provide service delivery in 
other areas of the state. By maintaining re- 
sponsibility for training staff members at other sites 
and by reviewing work performed at those sites, 
UTREP is working with DRS to give vocational 
rehabilitation clients across the state equal access to 
rehabilitation engineering services. 



Background 

In 1989, DRS approached UTREP to establish a 
contract for the provision of job accommodation 
services to DRS clients. The original agreement 
called for services to be performed only in west 
Tennessee. In each year since 1989, DRS has 
renewed its contract with UTREP and has worked 
with UTREP to improve service delivery to its 
clients. 

Approach 



Services are provided through a pre-paid contract, 
renewed yearly, known as the “Rehabilitation 
Engineering Services Project.” Providing services 
through a contract rather than through a fee-for- 
service arrangement benefits DRS counselors and 
clients because counselors do not have to wait for 
authorization before referring clients to project 
agencies. 



Because Memphis is located in the southwest comer 
of a state that is over 400 miles wide, UTREP could 
not reasonably provide rehabilitation engineering 
services for the entire state, even with the use of a 
mobile laboratory. Instead, UTREP and DRS 
selected the model of establishing at least one 
satellite site in each of the three grand divisions of 
the state (west, middle, and east). The initial satellite 
agency, the STAR Center in Jackson, Tennessee, 
was established in July 1990. Though Jackson also 



O 





lies in the west division of the state, staff members 
of the STAR Center provided the project with 
much-needed expertise in adaptive technology for 
persons with visual impairments. The East 
Tennessee Special Technology Access Center 
(ETST AC) joined the project as a satellite agency in 
July 1993 to serve the east division of the state. 

As the lead agency for the project, UTREP is 
responsible for training staff members at each 
satellite agency. Training includes topics such as the 
responsibilities of each project staff member under 
the contract, the operation of adaptive equipment, 
and evaluation and report-writing techniques. 
UTREP also provides technical expertise and 
performs product investigations for staff members as 
needed. On occasion, UTREP staff members may 
travel to other sites to assist in evaluations or other 
service provision. UTREP is also responsible for 
reviewing the evaluation reports from each site, 
ensuring that clients of each project agency receive 
the same recommendations as they would from any 
other project agency. 

The following services are currently available to 
DRS counselors and clients through the project: 

• Consultation. Counselors may refer clients for 
preliminary assessments to assist the counselor 
in determining if the client could benefit from 
pre-paid services available through the 
contract. 

• Evaluation. Counselors may refer clients for 
formal assessments for assistive technology. 
Each evaluation results in a formal report 
describing the evaluation process, the items 
recommended as a result of that evaluation, 
and a justification for the purchase of each of 
those items. The report also includes infor- 
mation on where each item can be purchased, 
and three or more vendors for each item for 
which bids must be received. The evaluator is 
always careful to avoid the presumption that 
DRS will purchase all or even any of the items 
recommended, because counselors often work 
with employers and with the clients themselves 
to determine responsibility for purchase. 

• Installation. Project staff members perform 
installation services when appropriate. 
Installation services are most commonly 
needed when a computer system has been 
purchased for a client. 



RESNA ’94 • June 17-22, 1994 



Engineering Service Provision for VR in Tennessee 



• Training. Project staff members train clients in 
the operation of any equipment, non-adaptive 
or adaptive, that they receive. The training 
may require only one session that takes place 
at the time of installation; however, when 
clients receive more sophisticated equipment 
(such as what is requir^ to allow the client to 
establish a home-based desktop publishing 
business), the training may involve repeated 
visits to the client’s home or place of 
employment. 

• Support. Project staff members provide 
support over the phone for questions clients 
may have regarding the operation of non- 
adaptive or adaptive technology that they have 
acquired through DRS. If problems cannot be 
resolved over the phone, project staff members 
will travel to the client’s home or place of 
employment over a period of weeks or months. 

• Lending library. A survey of DRS counselors 
in 1990 (1) showed that, after counselors 
received a report detailing recommendations 
for a computer system, the average delay 
before delivery of the system was 5'h months. 
DRS has since worked with UTREP to set up a 
lending library of computer systems, including 
adaptive access software and hardware, that 
can be provided to DRS clients as soon as the 
regional supervisor grants approval for the 
purchase of their systems. Tlie lending library 
is currently being expanded to include other 
types of technology, such as TDD devices and 
environmental control units. 

Each agency tries to be as consumer-responsive as 
possible. During each evaluation, clients are 
encouraged to share freely about preferences they 
may have for equipment or software, and about 
problems they may be encountering other than those 
that led to the referral for services by their coun- 
selor. When clients call regarding problems with 
equipment or software that cannot be solved over 
the telephone, staff members are almost always able 
to re-arrange their schedules to visit the clients on- 
site within one week. When DRS requested that 
evaluation reports be sent to the counselors as soon 
as possible, UTREP staff members redesigned the 
narrative portion of the evaluation report to reduce 
the time required to prepare each report. 

Discussion 

It is important for each client of the Division of 
Rehabilitation Services to have, whenever possible, 
equal access to assistive technology services to help 
them gain employment. By establishing sites around 
the state that provide consumer-responsive services 



as quickly and as uniformly as possible, DRS and 
UTREP have taken great strides toward achieving 
that goal. 

References 

1 . Cronk, Stan. 199 1 . A Survey of Tennessee DRS 
Employees on Assistive Technology. Proceed- 
ings of the RESNA Fourteenth Annual 
Conference. 199-201. 

Acknowledgments 

The author gratefully acknowledges the continued 
support of the Tennessee Division of Rehabilitation 
Services. 

Stanley R. Cronk 

Rehabilitation Engineering Program 
University of Tennessee, Memphis 
682 Court St. 

Memphis, TN 38163 
901/448-6479 




RESNA *94 • June 17 - 22 , 1994 



9 



KNOWLEDGE OF ASSISTIVE TECHNOLOGY IN SPINAL CORD-INJURED INDIVIDUALS IN 

KENTUCKIANA 



Robert P. Cunningham 
Department of Occupational Therapy 
Eastern Kentucky University 
Richmond, KY 



ABSTRACT 

Sustaining a spinal cord injury results in significant 
impairment in all areas of life. Specifically, 
dependency in the areas of activities of daily living, 
work/school and leisure results in the loss of 
control over the major functions of life. One 
approach the occupational therapist may use to 
address these areas is to provide assistive 
technology. The purpose of this study was to 
determine the knowledge of spinal cord-injured 
individuals residing in Kentuckiana regarding 
assistive technology. A mail survey of 84 persons 
with quadriplegia determined the tj^s of assistive 
technology the individuals were exposed to, 
whether evaluation was completed to determine 
device selection, was training provided in the use of 
obtained devices and who introduced these devices 
to the individual. The results indicated that the 
respondents expressed uncertainty/agreement when 
asked if they had received proper evaluation and 
disagreement and uncertainty when asked if they 
felt they had received adequate assistive technology 
services. The respondents were in agreement that 
they had been properly trained to utilize equipment 
they owned. 

BACKGROUND 

The treatment of the individual with SCI is a 
complex task. Regardless of the person's level of 
injury, the rehabilitation of the SCI individual will 
include the services of a variety of professionals. 
Areas that these professionals may address include 
training in activities of daily living, strengthening 
activities, communication, mobility, exploring 
avocational and vocational interests and skills and 
providing assistive technology to support these 
activities. Assistive technology is defined as "any 
piece of equipment, or product system, whether 
acquired commercially off the shelf, modified, or 
customized, that is used to increase, maintain, or 
improve functional capabilities of individuals with 
disabilities" (Technology-Related Assistance for 
Individuals With Disabilities Act of 1988). 

Due to the rapid development of new technologies, 
possessing adequate knowledge in the provision 
and use of assistive technology for persons with 
SCI is a difficult role for professionals to fulfill 
properly. While professional curricula are now 
addressing assistive technology, there remains a 
large number of professionals who may not have 




the knowledge to properly meet the role of 
providing assistive technology to persons with SCI. 

The purpose of this study was to conduct a survey 
of users of assistive technology devices to 
determine their knowledge regarding assistive 
technology, where they received this information 
and what training and services they received. 
Investigation of the problem served to identify 
those areas in which persons with SCI currently 
require the greatest assistance. 

RESEARCH QUESTIONS 

1. Are persons with quadriplegic SCI in 
Kentuckiana receiving adequate high tech assistive 
technology services? 

2. Are persons with quadriplegic SCI in 
Kentuckiana who utilize high tech assistive 
technology being evaluated properly for device 
selection? 

3. Are persons with quadriplegic SCI in 
Kentuckiana receiving adequate training in the use 
of devices they have received? 

4. ^o in Kentuckiana is making high tech 
assistive technology device recommendations? 

METHODOLOGY 

The research design used for this study was survey 
research. The information for the survey was 
gathered through the use of a mailed questionnaire. 
The instrument used to collect data was a 
questionnaire developed by the author. The 
questionnaire consisted of open-ended, closed- 
ended, contingency questions and attitudinal scales. 
The purpose of the questionnaire was to determine 
the types of assistive technology the individual has 
been exposed to, whether evaluation had been 
completed to determine device selection, was 
training provided in the use of obtained devices and 
who introduced these devices to the individual. 

The subjects included individuals with complete or 
incomplete quadriplegia who had received 
rehabilitation services at a rehabilitation hospital in 
Louisville, Kentucky. 

RESULTS 

Thuty-six or 51% of the 70 subjects returned the 
questionnaire. Demographic information revealed 



RESNA ’94 • June 17-22, 1994 



10 



Knowledge of Assistive Technology 

that the average age of the tespondents was 37.1 
years. The respondents level of it\jury varied from 
C3 to C7 with a mttjority (20%) of the respondents 
demonstrating an injury at the C6-7 level. The 
demographic information also revealed that 82% of 
the respondents indicated that they had not received 
information regarding funding for assistive 
technology devices. All of the 18% that had 
received funding information, had obtained this 
information bom a vocational counselor. 

Questions one through four asked the respondents 
to identify assistive technology devices Aey had 
tried out or owned and which rehabilitation 
professional introduced them to or recommended 
that they purchase each device. Devices the 
respondents had tried out are displayed in Table 1. 
Physical therapists (76%) were the professionals 
responsible for the majority of introductions to 
power wheelchairs with equipment vendors (24%) 
listed second. Occupational therapists, vocational 
counselors and others were equally identified 
(31%) as the professionals responsible for 
introduction to the computer. Fifty percent of the 
respondents who had tried out driving ai^ 
indicated that vocational counselors had inuoduced 
them to these devices. 



Table 1 

Assistive Technology Devices Tried Out 



Device 


n 


% 


Power Wheelchair 


25 


83 


Computer 


16 


53 


Driving Aids 


12 


40 


Environmental Control Unit 


3 


10 


Electric Page Turner 


1 


3 



Distribution of the devices owned by the 
respondents is presented in Table 2. Physical 
therapists (65%) were the professionals identified 
most frequently as being responsible for 
recommending purchase of the wheelchair, with 
equipment vendors (15%) and other (15%) being 
second. Vocational counselors (44%) and 
rehabilitation technologists (33%) were cited as 
those most responsible for recomendation of 
driving aids. Of the seven persons that had 
purchased a computer, five (71%) of the 
tespondents indicated that others were most 
responsible for recommending there purchase. 
Others included family, employer, psychologist and 
themselves. 

Table 2 also lists devices the subjects owned and 
were trained to use. Physical therapists (69%) were 
the professionals most identified as being 
responsible for providing wheelchair training. 
Vocational counselors (56%) were identified most 



frequently as the professional providing training in 
the use of driving aids. Other (66%) was identified 
most frequently as the professional responsible for 
the provision of training in the use of computers. 
Occupational therapists were most frequently 
identified for providing training in the use of 
environmental conuol units (66%) and electric page 
turners U00%) respectively. 

Table 2 

Numbers of Respondents Trained to Use Asaislfaa 

Devices They Q_wn 



Device 


Number 

Who 

Own 

Device 


Number 
Trained to 
Use Device 


% 

Trained 


Power 

Wheelchair 


20 


16 


80 


Driving Aids 


9 


9 


100 


Computers 


7 


3 


43 


Environmental 
Control Units 


3 


3 


100 


Electric Page 
Turner 


1 


1 


100 



The respondents were asked to identify the 
rehabilitation professional they felt had been the 
most helpful in providing assistive technology 
services. Physical therapists were identified as 
most helpful, occupational therapists were listed 
next in importance, and vocational counselors were 
listed third. Other and rehabilitation technologists 
were also mentioned. 

To ascertain the respondents views towards 
assistive technology and the services they had 
received, a 5-point Likert-type scale was used. One 
was strongly disagree and 5 was strongly agree. 
Results are shown in Table 3. 



Table 3 

Responses to Attitudinal Scale Ouestio PS 



Questions 


Mean 

Score 


I feel that I have a good understanding 
of how assistive technology can help 
me. 


3.6 


I feel I was properly trained to use all 
the assistive technology devices that I 
own. 


4.0 


I feel that I was properly evaluated for 
all the assistive technology devices 
that own. 


3.6 


Overall, I feel that I have received 
adequate assistive technology 
services. 


2.8 


Overall, I feel that assistive 
teeht{^ogv has been helpful to me. 


3.6 



RESNA ’94 • June 17-22, 1994 



11 



Knowledge of Assistive Technology 
DISCUSSION 

The results of the study are limited due to the 
respondents' diagnosis of quadriplegia, the small 
sample size, limited knowledge of £e respondents 
functional status and the fact that they represent a 
small g^graphical area. Because of these factors, 
the ability to genmlize the findings beyond 
Kentuckiana or according to level of injury is not 
possible. 

The large number of respondents who indicated 
that Aey had not receiv^ information regarding 
funding indicates a strong need for rehabilitation 
professionals to provide this information. Also, the 
fact that the vocational counselor was the only 
professional cited as providing funding information 
indicates a need on the part of all professionals to 
become more knowledgeable and active in this 
area. 

The power wheelchair was identified most 
frequently as the assistive technology device that 
the respondents had tried out, owned and been 
trained to use. This information is not surprising in 
that the wheelchair serves as the primary source of 
mobility, postural support and is fiequfently covered 
by insurance. The fact that the physical tho^ist 
was identified as the person most responsible for 
introducing, recommending purchase and providing 
training for this device may serve as the reason why 
the physical therrqrist was viewed as the 
professional most helpful in providing assistive 
technology services. 

The compute was the second most fiequently 
identified device that the respondents had tried. 

Less than one-fourth of the respondents indicated 
that they owned a computer. Related to this was 
the finding that others were the persons most cited 
for recommending purchase. This suggests that 
computers are being used as treatment tools, but are 
not being pursued as a possible means for increased 
independence. 

Physical therapists, occupational therrqrists and 
vocational counselors were the professionals listed 
as being most helpful in providing assistive 
technology services and as the professionals 
responsible for recommending devices and 
providing device training. This indicates that these 
professionals need to make a concerted effort to 
maintain or increase their knowledge in the area of 
assistive technology services. The fact that 
rehabilitation technologists were seldom listed 
indicates a need for increased involvement by this 
professional group in the provision of assistive 
technology services. 

The most striking aspect of survey findings were 
the results of questions utilizing the Likert-type 



scale. The 2.8 mean score in response to the 
statement asking if they had received adequate 
assistive technology services indicates that they 
perceive that they have not received adequate 
services. This information coupled with the 3.6 
mew scores in response to the statements regarding 
having an understanding of assistive technology, 
the belief that they received prtqrer evaluation and 
the belief that assistive technology has been helpful 
to them suggests a need for improving the provision 
of assistive technology services. It may also 
indicate that more thorough patient education is 
needed in regards to what assistive technology may 
be beneficial or not practical fw the potential users. 
It is noted that the dka strongly suggest that proper 
training is being provided for those devices which 
the respondents receive. 

Reviewing the research questions of this study 
indicates that assistive technology services 
povided to persons with SCI in Kentuckiana could 
be improved. The study also suggests that persons 
with SCI in Kentuckiana are being properly 
evaluated for and trained in the use of assistive 
technology devices. Finally, physical therapists, 
occupational therapists and vocational 
rehabilitation counselors are the persons most 
responsible for assistive technology 
recommendations. 

It is recommended that this study be replicated for 
the entire state of Kentucky in order to generalize 
the findings. Further research is suggested to 
identify the needs of all persons who utilize or may 
benefit from assistive technology. 

ACKNOWLEDGEMENTS 

Subjects for this project were provided by Frazier 
Rehabilitation Center in Louisville, Kentucky. 
Thanks is offCTed to Frazier’s staff and their Spinal 
Cord Task Force for all the assistance they 
provided. 

REFERENCES 

Technology-related assistiance for individuals with 
Disabilities Act of 1988. (P.L. 100-407). 

Robert Cunningham MS, OTR/L 
Instuctor 

Eastern Kentucky University 
Departoent of Occupational Therapy 
103 Dizney Building 
Richmond, KY 40475 
((606)624-3300 



26 



12 



RESNA’94 • June 17-22, 1994 



PROJECT IMPACT: DEVELOPMENT OF A STATEWIDE SERVICE DELIVERY MODEL 
through local CAPACITY BUILDING 



Lawrence H. Trachtman' and Maggie Sauer® 

‘N.C. Assistive Technology Project, Division of Vocational Rehabilitation Services, Raleigh, NC 
® Center for Development and Learning, University of North CaroUna at Chapel Hill. Chapel Hill. NC 



ABSTRACT 

Project IMPACT represents a partnership among three 
organizations to improve and expand access to assisti ve 
technology services for North Carolinians with 
developmental disabilities. Theproject’soverall purpose 
is to build assistive technology service delivery capacity 
in North Carolina. IMPACT operates under the following 
four goals: 1 ) develop a comprehensive array of assistive 
technology services by linking existing programs, 2) 
expand assistive technology service delivery capacity 
statewide, 3) provide training to consumers, families 
and professionals on assistive technology use, and 4) 
establish networks for information exchange and support 
among families, consumers and professionals. Mter 
three years, IMPACT will have increased service 
capacity and will have demonstrated a statewide service 
delivery model making assistive technology more 
available and accessible to individuals with 
developmental disabilities. This paper will present the 
model and the first year’s results. 



Certainly there is no one model for statewide delivery of 
assistive technology services. Most, however, would 
agree tlmt state planning should be based upon needs, 
available resources and a set of realistic goals (^. In 
addition, planning must occur locally to meet citizens’ 
needs and should take into consideration emerging 
critical issues such as cost/beneHt, quality assurance, 
outcome measurement and trainin^certification. A« 
interesting approach, and one which this project is 
based upon, is described by Schoech, Cavalier and 
Hoover (8). The Community Assistive 'Technology 
Services Network (CATSN) is a model for integrating 
technology into a multi-agency community human 
services delivery system. Although the development 
and testing of the model focused on systems change 
rather than on the creation of new services, the 
arrangement of service modules has clear implications 
for a statewide service delivery system. The authors 
present some of the barriers they faced during the 
project, but conclude that implementation of the model 
can create a coordinated network of technology services. 



BACKGROUND 



APPROACH 




Statewide planning for delivery of assistive technology 
services is a relatively new phenomenon. It was not 
until the few years prior to passage of the Technology- 
Related Assistance forindividuals with Disabilities Act 
of 1988 (P.L. 100-407) that there was even much 
interest in coordinated statewide service systeins. Two 
of the early planning efforts took place in Minnesota 
and New York where Govemor-s^jpointed task fon»s 
studied consumer needs and made recommendations 
for improved technology services (1,2). After the Tech 
Act was passed, other states began developing their 
grant applications around statewide services provision. 
Parette and VanBiervliet, in a report on the needs of 
infants and young children with disabilities in Arkansas, 
recommend a decentralized system of servic^ to account 
for problems with transportation, information, funding 
and lack of hands on technology experiences by families 
living in rural areas (3). In South C^olina, Langton and 
Trachtman propose a cooperative service delivery model 
in which agencies operate within a regional service area 
to share staff and resources for more effective technology 
services (4). The results of the model’s first year of 
operation suggest it is feasible if the benefits outweigh 
the costs of participating and if a facilitator can help 
maintain the necessary relationships (5). And in yet 
another approach, Kniskern proposes a systems 
integration model in which three systems levels (policy 
makers, service providers and consumer groups) interact 
both vertically and horizontally resulting in increased 
service capacity and a better match between consumer 
needs and services (6). 



Project IMPACT (Innovative Methods for Providing 
ACcess to Technology) responds to an identified n^ 
in North Carolina for increased access to assistive 
technology devices and services for individuals with 
developmental disabilities and their families. As 
identified in the Technology-RelatedNeeds Assessment 
Project - Final Report (9), consumers and families need 
more information on how to obtain and repair equipment, 
as well as appropriate training on how to use equipment 
Service needs include technology-related consultation, 
equipment demonstration and equipment loan/rental 
and repair. Professionals want more preservice and 
inservice training on assistive technology and see a 
significant need for follow-along services. Consumers 
and professionals agree that consumers need to incre^ 
their awareness level regarding available assistive 
technology. Project IMPACT is a Developmental 
Disabilities Council-funded grant designed to address 
these needs through a comprehensive three-year plan of 
activities. 

Project IMPACT has developed a model to increase 
service delivery capacity in North Carolina (figure 1). 
The principal organizations are the N.C. Assistive 
Technology Project (NCATP), the Clinical Center for 
the Study of Development and Learning (CDL) - a 
University Affiliated Program and the Family Support 
Network of N.C. (FSN). Linkages and grant management 
are provided by NCATP acting as a central coordinating 
hub. The following four goals expand the concepts 
Shown in the figure. 




RESNA ’94 • June 17-22, 1994 



13 



PROJECT IMPACT 



Prelect IMPACT 

Innovative Methods (or Providing ACcess to Technology 



Center tor Dsyetopment & Learnina 




Family Support Network 


Irnitiliig 




tiiroriiinlloit & Rcfcrml 


mitl 


. • • • 


niid 


fCilucnltoii 




FJocIroiilc Nchvorkliiq 




N.C. Asslsttvfl lochnotopy Pro|BCt 



Oiitnil 

CoordiiinHoii 



Center lor PovotopmBnl ft LBarntno 



Client 

Services 



Figure 1 

Goal 1. Link existing programs 
A project management team consisting of the project 
dilator, coordinator andrepresentatives from CDL and 
FSN meets monthly to plan and review all activities. A 
nine-member interagency advisory board helps guide 
the project. The management team works with the 
advisory board to develop service linkages and 
interagency partnerships, including developing a client 
tracking and evaluation system which could be used 
among all agencies providing assistive technology 
services. As the project progresses, the management 
team will be responsible for idendfying barriers to 
acquiring assistive technology (such as funding, lack of 
trained professionals, limit^ access to device try-out) 
and will develop clear and concise plans for policy and 
systems changes. The management team works closely 
with the NCATP to help implement these changes. In 
addition, the project director and the management team 
^ responsible for developing a plan for continuing 
project activities beyond the three-year grant 

Goal 2. Expand service delivery capacity 
A multidisciplinary, state training team helps with 
curriculum development, training activities and regional 
consultation/coordination among new and existing 
programs. Four regional teams will be established 
whose role is to provide multidisciplinary assistive 
technology assessments/follow-along services and also 
to help build regional resource networks. In addition, 
local programs are selected in each of the four regions 
to serve as a model, community-based pilot projects. 
Pilot projects demonstrate exemplary practices of 
technology utilization across select ages groups. 
Regional teams and pilot projects work closely with 
local resources, such as the NCATP’s technology 
demonstration centers, to esthblish a regional service 



delivery network. CDL provides intensive, ongoing 
consultation to both the regional teams and pilot projects. 
Pilot projects and regional teams then help with 
replication following their year of receiving technical 
assistance. A technology demonstration center in 
Wilmington expands service capacity in that part of the 
state. 

Goal 3. Provide_fraining to consumers, families and 
professionals 

A comprehensive assistive technology training 
curriculum based upon existing training materials will 
be developed, test^, revised and completed by the 
project’s end. This curriculum is used to help train the 
regional assistive technology teams and pilot projects. 
This training is provided by the state training team 
members who work closely with CDL to provide 
technical assistance and ongoing support In addition, 
learning modules which have a consumer and family 
focus will be developed. Learning modules may cover 
such topics as technology and literacy, technology for 
learning disabilities and technology for the elderly. The 
learning modules will be disseminated to interested 
organizations. 

Goal 4. Establish networks for information exchange 
North Carolina, through FSN, currently offers a toll- 
firee assistive technology information and referral s^vice. 
Another key component for information exchange is an 
electronic computer bulletin board service. This bulletin 
board is available to consumers, families and 
professionals statewide through computer/modem 
hookup. The bulletin board offers electronic mail, 
discussion forums and provides current information on 
such topics as statewide training sessions, used 
equipment, legislation, and questions and answers on 
problems related to technology applications. In 
conjunction with the bulletin Ixwd, a peer support 
network among consumers and families helps match 
experienced assistive technology users (mentors) with 
new technology users (peers). 

When fully implemented. Project IMPACT will have 
increased service capacity by developing and 
demonstrating a model, statewide assistive technology 
service delivery system. Model components, replica^ 
across four regions of the state, will consist of a regional 
assessment team, a technology demonstration center, 
and a local pilot project. A statewide information 
network featuring toll-free and electronic access will 
facilitate communication into and among the service 
components. Training materials will help consumers 
and families become more knowledgeable of assistive 
technology, and a peer matching system will facilitate 
support among individuals and families who have similar 
ne^s. 

RESULTS 

Project IMPACT began operation in January 1993. The 
management team, advisory board and state training 
team have been integral to the project’s operation. A 
regional team and pilot project were established in the 
north central region through competitive applications. 



14 



28 



RESNA ’94 



June 17-22, 1994 



PROJECT IMPACT 

The regional team is located at Murdoch Center, a 
residential facility for individuals with mental retardauon 
and developmental disabilities. The Murdc^h Center 
team will provide comprehensive, multidisciplinary 
team assessment and follow-along services to residents 
and consumers in their 16 county region. Thej^ill also 
provide training and community outreach. The pilot 
project is located at the Bowman Gray School of 
Medicine. Through their Physicians’s Home Cwe 
Program, this model project will help older pwple who 
live at home remain independent duough the use ot 
technology. A comprehensive training cumculum 
responsive to the programs’ needs was prepared by 
CDL with input from the state team, and two mulu-day 
training workshops were conducted. CDL has continued 
to work extensively with Murdoch Center and Bowman 
Gray and is helping them develop short’ and long’tenn 
program goals. A data collection system is in the fiijm 
stages of development and will be implement^ with 
theprogramsbeginning in year two. Andin Wilmington, 
the New Hanover Regional Medical Center has been 
selected for the new technology demonstration center 
site. 

The biggest challenges faced by IMPACT have been in 
implementing the statewide electronic bulletin board 
and the peer matching program. Issues with personnel 
and hardware/software forced significant changes 
midyear to the bulletin board. However, a r^nt 
development appears very promising. The Raleigh 
News and Observer (the state’s largest newspaper) has 
proposed a partnership whereby people with disabilities 
will be given free access to their statewide computer 
bulletin board service. Features will include on-line 
publications (such as the newspaper), discussion forums, 
electronic mail, Internet access, home shopping and 
games. A person with a disability will serve as systenis 
operator. A written agreement is expected earty m 
1994. The peer matching program also has b^n 
restructured and that too is expected to begin operation 
this year. 

DISCUSSION 

Project IMPACT has been a major undertaking, both in 
scope and management. The proposed service delivery 
model has been well received and all involved P»iues 
have contributed significantly to its operation. Much of 
the support has been volunteer, including the state 
training team and the Murdoch Center and Bowman 
Gray programs. However, we feel that both programs 
have “bought-in” to the model and are committed to its 
success. The service network is now in place in one 
region of the state and the information and peer support 
overlays will facilitate communication into and among 
the components. By linking new and existing resources, 
we expect to build strong regional services thereby 
creating an even stronger statewide service delivery 
network. Data collection and outcome measurement 
will help validate the model’s key attributes. 

Unfortunately, a budget cut back has forc^ a reduced 
effort for year two. However, the service delivery 



RESNA’94 • 



1 . 

2 . 

3. 



model remains viable and local capacity buUding will 
be pursued. Project IMPACT’S plans for future support 
consist of 1) continuation and expmsion of the Hi^ 
sponsoring programs. 2) formation of a statewide 
p^ership among agencies serving individuals with 
disabilities. 3) maintaining r^ional and pilot 

programs initiatedby Project IMPACT through fee-fOT’ 

services, fundraising andin-ldnd support tuid4) assisting 
regional programs to work together for more effective 
and efficient service delivery. Our goal is that ^e locd 
and statewide resources developed by the project will 
lead topublicand private support for these new services, 

REFERENCES 

Governor' s Report on Technology for People witii 
Disabilities. State of Minnesota, June 1986, 

A Final Report of the Task Force on Technology 
and DisabUities, New York State, October 1981 
Parette, H.P. Jr. and VanBiervliet (1991). 
Rehabilitation Assistive Technology Issues for 
Infants and Young Children with DisabUities: A 
Preliminary Examination. Journal of Rehabilitation, 
57f3)’ 27-36, 

Langton. A.J. and Trachtman. L.H, (1989). 
Assessing the AvallabUity of Program Resources 
and Future Nee^ within a State. In l^ovision of 
Assistive Technology, Planning and 
Implementation, Report of a workshop hosted by 
the Electronic Industries Foundation, Washington, 
DC 

Caster, L.S., Langton, AJ. and Trachtman, L.H. 
(1991). Cooperative Service Delivery: A Cost- 
Effective Strategy. In Proceedings of the 14th 

Annual RESNA Conference, Kansas City, MO. 

Kniskem.J. (1991). A Systems Integration Model 
for Assistive Technology Resource Development. 
In Proceedings ofTouch the Future, Third Southeast 
Regional Conference on Assistive Technology. 
Atlanta, GA. 

Enders.A. (1987). Planning for and Implementmg 
Rehabilitation Technology Services. Electronic 
Industiies Foundation, Washington, DC 
Schoech, D., Cavalier, A. and Hoover, B, (1993). 
A Model for Integrating Technology into a Muln- 
Agency Community Service Delivery System. 
Assistive Technology, 5(1): 11-23. 

. Technology-Related Needs Assessment Project 
Fin^ Report, Prepared for the N,C. Council on 
Developmental Disabilities, April 1991. 

tCKNOWLEOGEMENT 

Toject IMPACT is supported in part hy^*6 North 
:wolina Council on DevelOTmen^ Debilities wd 



Lawrence H. Trachtman 

North Carolina Assistive Technology Project 
1110 Navaho Drive, Suite 101 
Raleigh, NC 27609 

23 

lune 17-22. 1994 . 



15 



REHABOL.ITATION TECHNOLOGY SERVICE DELIVERY MODELS IN 
VOCATIONAL REHABILITATION AGENCIES: A MULTI-LEVEL APPROACH 



Cyntbia C. Flynn, Pb.D. 

Center for Rebabilitation Technology Services 
South Carolina Vocational Rehabilitation Department 



Abstra0 

Current methods for classifying r^abilitation 
technology service delivery models do not 
adequately describe these services in Vocational 
Rebabilitation (VR) agencies. The purpose of this 
paper is to present, an alternative classification 
scheme. This scheme is being used in a national 
study to evaluate rehabilitation technology services 
in VR programs. 

Background 

Vocational Rehabilitation agencies, funded by the 
United States government, are required to provide 
r^abilitation technology services. A preliminary 
survey of VR agencies (CRTS, 1992) reported 
that they do provide some rehabilitation 
technology services to their clients. There were 
significant differences in the way these services 
were delivered and bow these programs were 
managed. VR agencies are looking for ways to 
deliver these services in the most efficient and 
cost effective manner possible. In order to 
conqrare service delivery programs between 
agencies, some pattern, model, or other common 
basis must be identified. One method typically 
used in describing and analyzing service delivery 
programs is to categorize them according to some 
characteristic or set of characteristics. This paper 
focuses on classifying service delivery tgiproaches 
used in VR agencies in order to then evaluate 
efficacy of the various models. 

The Thirteenth Institute on Rehabilitation Issues 
(Cortbell & Thayer, 1986) reviewed models of 
rehabilitation technology service delivery in a 
variety of settings including VR agencies, In their 
review, they identified models, however, no 
classification system was used to distinguish 
among the various programs nor was my external 
evduation performed. While some of these 
models continue today and some programs have 
been replicated, most service delivery programs in 
VR agencies have not been developed following 
these or any other model. 

Smith (1987) identified seven descriptive factors 



which are frequently used to classify and describe 
rebabilitation technology service delivery models; 
(1) purpose and mission; (2) functional area; (3) 
geographical catchment area; (4) population; (5) 
internal operations; (6) location; and (7) 
administrative home base. A closer examination 
of these descriptive factors reveals that, while 
they may be useful in a general sense, many are 
not useful in discriminating among rehabilitation 
technology service programs in VR agencies. The 
descriptive factor "administrative home base" 
refers to the type of orgmizational setting in 
which the program operates. Since all of the 
programs to be studied are part of VR agencies, 
this factor would not discriminate among 
programs. Several other descriptive factors - 
purpose md mission, population, and functional 
areas - would not adequately discriminate VR 
rebabilitation technology service delivery 
programs because, as a part of the VR program, 
these programs must follow federal requirements 
which mandate that rehabilitation technology 
services be available to all eligible clients. They 
could differ in hss they provide these services 
but not whether , or to whom, they provide them. 
Factors wUcb are useful in discriminating among 
VR programs include: geographical catchment 
area, internal operations, and location. 

"Geographical catchment area" is the geographical 
territory which the program serves, typically 
state, county, or community. This is viewed as 
one of the most common ways to categorize 
service delivery programs (Smith, 1987). 
Although all VR agencies provide statewide 
services, the way in which they organize the 
delivery of rehabilitation technology services 
varies across agencies. For example, some 
agencies divide the state into multi-county regions 
and assign technology staff to deliver services in 
a specified region. Other agencies do not divide 
the state. Instead, rebabilitation technology 
services are provided from an office or several 
offices each of which serves the entire state. 

"Internal operations" addresses the management, 
staffing pattern, structure, or other aspect of 
program operations. One aspect of staffing 



16 



30peSNA ’94 



June 17-22, 1994 



Service Delivery Models in VR 



pattern relates to type of staff. A VR program 
might follow the single expert model) the 
interdisciplinary team model, or some model in 
between. Staffing can also vary from "employee- 
based” with full time technology staff hired by the 
agency to "outside provider-based" with 
contractual-based service providers. 

"Location” refers to the position, station, or 
situation in which services are provided. Location 
discriminates among VR rehabilitation programs. 
Langton, Caster, and Trachtman (1991) identified 
two locations: stationary and mobile. A mobile 
service delivery unit operates out of a vehicle that 
travels to see clirats in their homes, jobs, or at a 
convmient public facility such as a regional VR 
office. The primary purpose of the vdiicle is to 
provide a place to conduct assessment, 
fabrication, and/or other services. The stationary 
service delivery approach operates out of a 
facility. Clients come to the facility to receive 
rehabilitation technology services. The service 
provider may occasionally go out to other sites to 
provide assessment or delivery of a device, etc., 
but the base of services is non-mobile. Using this 
concept of stationary, a more accurate term might 
be facility-based. This may seem confusing since 
most people who provide rdbabilitation technology 
services go to the consumer’s home, work place, 
the local VR office, or some other facility. The 
difference is that the mobile unit carries with them 
the capability to fabricate, repair, or provide 
related services on site without going back to their 
home office. 

Statement of the Problem 

• A classification scheme is needed which 
discriminates among VR systems in order 
to compare efficacy of the various 
programs that exist. 

• Current classifications do not, in their 
entirety, discriminate among the various 
VR rehabilitation technology service 
delivery programs. 

• Three factors from previous classification 
schemes do appear to discriminate, 

• Since all of these factors discriminate 
among VR rehabilitation technology 
service delivery programs, using all of 
these factors will provide a more 



powerful classification of programs. 
Approach 

Eight service delivery models are proposed to 
classify VR rehabilitation technology service 
delivery programs using the three factors 
identified as discriminators. In order to classify 
programs into one of the above models, several 
questions were developed for a comprehensive 
survey sent to ail VR agencies. Questions 
included: how the delivery of rehabilitation 

technology services are structured, what 
percwtage of r^iabilitation technology services 
(apart from product sales) provided to VR clients 
are deliver^ by VR employees and what 
percentage are provided by outside service 
providers, and what percentage of rehabilitation 
technology services offered by VR are provided 
from a vehicle and what percentage from a 
facility. 

Results and Discussion 

Partially or fully completed surveys were returned 
from 62 VR ag^cies for a return rate of 11%, 
Of these, 58 conqjleted Section 3 of the survey 
which dealt with rehabilitation technology 
services. Where the program does not fall at one 
end of the continuum or the other (e.g. , they have 
employees and they contract with outside service 
providers), the predominant characteristic is used 
to classify the program. For example, if 75% of 
their services are provided by employees, they are 
classified as internal. 

The models reported by states are: 

regional/intemal/facility-based [N=9] 
central/internal/facility-based [N^13] 
regional/extemal/facility-based [N^ 19] 
central/extemal/facility-based IN=8] 
regional/intemal/vebiele-based [N-1] 
central/intemal/vehicle'based [N=0] 
regional/extemal/vehicle'based [N=l] 
eentral/§xtemsl/v6hicl§"based [N=2] 

This system appear to discriminate among 
rehabilitation technology service delivery 
programs in VR agencies. As with any 
classification system which is imposed after the 
program was developed, it doesn’t totally fit, 
One difficulty with the descriptive factors 
proposed is that they exist on a continuum. For 



RESNA’94 • June 17-22, . 



17 



Service Doliveiy Models in VR 



example, the internal operations can exist on a 
range from 100% employees to 100% outside 
service providers. The n^el focnges on the ends 
of the continuum for each factor, In most cases, 
the program existed closer to one end of the 
continuum. For those programs that existed 
closer to the middle, classification was more 
difficult and less meaningful. Some programs 
even had both regional and cmitralixed services 
operating within their agency. This usually 
occurred with some specialized program which 
provided »rvices across the state while routine 
rehabilitation technology services were provided 
by region. In these cases, the program was 
classified as regional since most of their services 
were provided using the regional model, 
ClassiFcation according to internal operations 
appears to be closer to the middle of the 
continuum rather than at one end or the other, 
Five programs could not be identified as eittter 
internal or external because they have 50% 
en^loyees and 50 % outside service providers. Of 
the five programs which could not be classified 
according to intemal/extemal, only one was 
vehicle^based. Very few agencies (5) pripiarily 
provide these services using the vehicle-based 
model. Currmjt work uses this classification 
system to analyze the efficacy of the various 
rehabilitation technology services delivered in VR. 
Detailed data on cost, activities, and time use are 
underway in pursuit of this goal. 

Rrferences 

Corthell, D.W., & Thayer, T. (1986). 
Rehabilitation technologies. Thirteenth 
Insti tute on Rehabilitation Issues . 

Menomonie, WI: University of Wisconsin- 
Stout. 

<’RTS. (1992). Vocationd rehabiljtatiori: 
Assistive technolo gy survey research focus 

elements. (Available from Center for 
R^abilitation Technology Services, Post 
Offlee iox 15, West Columbia, SC 29171) 

Sndtb, R.O, (1987), Models of service delivery 
in rehabilitation technology. In L.g, 

Perlman & A, Bnders (ids,), Rehabilitation 
technology s ervice delivery: A nfartip^al 

gui^ ^p. 9-25). Washington, DC: l^SNA 
Press. 



Acknowledgments 

The Center for Rehabilitation Technology 
Services (CRTS) is part of the South Carolina 
Vocational Rdiabilitation Department, Support 
for this work has been provided through the 
National Institute on Disability and 
Rdiabllltation Research (NIDRR), U.S. 
Department of Education, Washington D.C. as 
part of the r^bilitation engineering research 
center grant #Hl33E20002-93. 

Cynthia C. Flynn 

Center for R^abilitation Technology Services 

South Carolina Vocational Rehabilitation 

Department 

1410-C Boston Avenue 

W. Columbia, SC 29171 



C 



18 



RESNA ’94 • June 17-22, 1994 



TETRA society of north AMERICA: THE FIRST THREE YEARS 



SAM SULLIVAN AND BRIAN GIFFEN 
TETRA SOCIETY 



ABSTRACT 

The range of physical 
disabilities is enormous, 
and the needs of people 
with the same disability 
varies greatly with each 
individual; thus it is 
impossible for the market 
to provide all the 
products and assistive 
devices people with disa- 
bilities require. Tetra 
Society of North America 
has united technical vol- 
unteers and disabled 
people to provide hundreds 
of low-cost assistive 
devices from 1990-1993. 



BACKGROUND 

Organizations have been 
formed to mobilize skilled 
volunteer labour to provide 
the technical assistance 
people with disabilities 
require. In Australia a 
highly successful program 
known as Technical Aid to 
the Disabled (TAD) matches 
professional engineers on a 
one-to-one basis with 
clients with disabilities. 

A similar program called 
REMAP operates in Britain. 

In Vancouver, Canada, Tetra 
has developed the Technical 
Assistance Program (TAP) 
modeled after TAD and REMAP . 



The organization was founded 
and is directed by a quadri- 
plegic, Sam Sullivan. Tetra 
has completed over 500 pro- 
jects since its inception in 
1989. The group has over 250 
active volunteers and 15 
chapters operating across 
Canada. Tetra is presently 
undertaking expansion by 
development of additional 
chapters across North America. 

OBJECTIVE 

Tetra has two objectives: 

* to assist people with 
severe disabilities to become 
as independent as possible so 
they can fully participate in 
their own communities 

* to promote awareness of 
the abilities, skills and 
talents of people with disa- 
bilities and to encourage 
their integration into the 
workplace 

DISCUSSION 

An organization linking 
volunteers with disabled 
individuals to provide low- 
cost assistive solutions 
provides several community 
benefits : 

* customization. Tetra is 
able to offer the disabled 
client complete customization 
of the device through direct 
linkage with the technical 
volunteer 




RESNA’94 • June 17-22, 1994 



19 



TETRA SOCIETY OF N.A, - THE FIRST 3 YEARS 



* cost effectiveness. The 
devices are provided in a 
manner that is far less 
expensive than similar 
solutions produced through 
institutional channels, 
because of the volunteer 
component of the production 
process 

* promotion of inclusion. 
The devices produced allow 
disabled people to enter 
into daily activities 
previously barred to them, 
promoting inclusion and 
independence 

* promotion of sensitivity 
and awareness. Volunteers 
become sensitized to the 
needs of the physically 
disabled and the potential 
means by which these needs 
can be met. Medical pro- 
fessionals working with 
Tetra learn from the 
volunteers how simple and 
inexpensive many technical 
solutions are : 

ACKNOWLEDGEMENT 

Tetra receives funding from 
a variety of community, 
corporate and governmental 
groups . 




Mr. Sam Sullivan 
Executive Director 
Second Stage Rehabilitation 
Society 

Suite 27, 770 Pacific Blvd. 

South - Plaza of Nations 
Vancouver, BC V6B 5E7 
phone (604) 688-6464 
fax (604) 688-6463 



RESNA ’94 • 



20 



June 17-22, 1994 



PROJECT IMPACT: Innovative Methods for Providing Access to Technology 



Maggie Sauer (1), Lawrence H. Trachtman (2), and Debbie Reinhartsen(l) 

\ Center for Development and Learning, University of North Carolina at Chapel Hill, Chapel Hill, NC 
2N.C. Assistive Technology Project, Division of Vocational Rehabilitation Services, Raleigh, NC 



ABSTRACT 

Project IMPACT represents a partnership 
among three organizations to improve and 
expand access to assistive technology 
services for North Carolinians with 
developmental disabilties. A large 
component of the project is devoted to the 
selection and training of regional assistive 
technology teams to expand service 
delivery capacity. Additionally, pilot 
projects which showcase the unique 
application of assistive technology are 
selected and included in the resources for 
each region. The regional assistive 
technology teams and pilot projects are 
provided with training on a variety of 
topics to enable them to address evaluation, 
follow-up and specific project related 
training needs. This paper will present the 
training model used and actviiies from the 
first year. 



BACKGROUND 

As the theoretical and research base for assistive 
technology service continues to expand, so does 
the need for effective service delivery and training 
models. A variety of service/training delivery 
models have been proposed to best provide 
assistive technology services (Trachtman & Sauer, 
1994; Beukelman D., & Mirenda P, 1992; 
Schoech, D., Cavalier, A., and Hoover, B., 1993; 
Blackstone, S., 1989; Blackstone, S. & Casselt- 
James, E. 1988; Matas, J., Mathy-Laikko, P., 
Beukelman, D., and Legresely, K. 1985; Cohen, 
C. & Frumkin, J., 1988; Cohen, C. 

1986;Vanderhaiden, G. & Lloyd, L. 1987; Shane 
H., & Yoder, D. 1981). Often, consumers and 
families travel great distances to receive services. 
Once the service is provided, follow-up and long 
term management are difficult for all concerned. 
Project IMPACT represents a compilation of 
various attributes of these described service 
delivery models as a vehicle for meeting 
service/training needs in North Carolina. The 



mechanism for accomplishing these tasks was 
created by utilizing and expanding the 
capabilities of existing programs in the state 
(Center for Development and Learning- 
University of North Carolina, North Carolina 
Assistive Technology Project, and Family 
Support Network- University of North Carolina). 
The project activities were selected based on 
needs surveys conducted slate-wide as well as 
those described by parents, consumers and 
professionals. It is the training component of the 
project which this discussion will describe. 

A variety of service delivery and training issues 
have been described by the above authors. 

These include variables related to specific 
disciplines as well as a host of broader issues 
such as administrative policy, client 
programming, professional training, referral 
process, resource availability, staff roles, 
interagency communication and cooperation, 
information and program awareness, technology 
availability, funding programs and equipment. 
As described by Church and Glennen (1992), ” a 
newly developing assistive technology program 
should seek to recognize any constraining factors 
and to develop the best possible program within 
those limitations. Rather than trying to develop 
an all encompassing assistive technology 
program that tries to do all things for all persons, 
the team should limit services to those areas 
which can be performed well." It is this 
approach which was adopted to address the 
training initiatives of Project IMPACT. 

APPROACH 

Project IMPACT divides the state of North 
Carolina into four regions. Each of the these 
regions is supported by a regional resource center 
where parents, consumers and professionals can 
come for additional information, hands-on 
experience with devices (lo-tech and high tech). 
Each offers a varying array of educational 
programs from funding to specific device 
training and use. Evaluation and service delivery 
is not offered through the centers. In an effort to 
provide a service delivery system and link 



35 

RESNA’94 • June 17-22, 1994 



21 



PROJECT IMPACT 



parents and consumers state-wide, the project 
proposed a service delivery system which would 
offer the following components to each region of 
the state: 

a) interdisciplinary service delivery team for 
evaluation, intervention and follow-up; b) pilot 
project to showcase special applications of 
assistive technology and augmentative 
communication strategies; c) linkage of service 
delivery teams and pilot projects to regional center 
activities to provide resource information and 
hands-on for parents, consumers and 
professionals; d) electronic bulletin board to link 
parents and consumers directly. 

Training Model 

Each of these components has the potential of 
enhancing and expanding service delivery 
capacity statewide. However, as recognized by 
Cohen (1986), "technology teams require highly 
trained professionals who specialize in the field." 
Additionally, as addressed by Church and 
Glennen (1992), "a multidisciplinary model of 
providing assistive technology services requires 
coordination across disciplines. A team leader is 
needed to coordinate services and resolve team 
differences. The team leader needs to be familiar 
with technology issues across all team areas but 
does not need to be an expert in each discipline." 

In order to expand services it was necessary to; 1) 
identify existing resources within each region of 
the state, 2) identify the assistive technology and 
pilot project teams and their individual and 
organizational capabilities, and 3) provide 
training and follow-up support to the teams 
identified within each region. Training activities 
were provided utilizing expertise from 
professionals, consumers and family members 
statewide to regional and local programs. 

A multidisciplinary state training team was 
established to recognize existing assistive 
technology expertise in each region and to 
furthermore provide representation from each 
region of the the state. Their involvement 
includes recommendations for specific activities 
within each of the four regions of the state and 
curriculum and material development. 

Regional assessment teams and pilot projects are 
selected in each region on a competitive basis. To 
date, two regional assessment teams have be^ ^ 



selected and one pilot project. The pilot project is 
a home health care program serving persons who 
are elderly to enable them to stay in their homes 
rather than moving to an institutionalized setting. 
These groups will maintain an ongoing 
relationship with Project IMPACT. The initial 
training phase is approximately nine months long 
followed by follow-up and consultation to the 
teams as needed. This is determined largely by 
team development and differs from group to 
group. Factors which affect this include 1) 
familiarity with assistive technology at the onset of 
project activities 2) organizational capability of 
the team 3) organizational capability of the agency 
4) requests by the team for assistance. 

Traininfl Activities 

Following selection, each group was given a 
questionnaire which closely mirrored the proposed 
training curriculum. Team members were asked 
to identify their level of expertise across each of 
the curriculum topics. These questionnaires were 
used by the training staff to customize training to 
the identified strengths and weaknesses of the 
team as a whole and individuals. 

Training included: 1) two, three-day training 
workshops at the beginning and midway through 
the training period, 2) ongoing on-site 
consultation weekly or bi-weekly, 3) ongoing 
consultation via electronic link directly to training 
staff at the Center for Development and Learning 
4) observation of team evaluations and service at 
the Center for Development and Learning 5) 
working closely with regional assistive technology 
resource centers to network goals of selected 
programs with the center. 

The training curriculum includes a compilation of 
articles, resource lists and excerpts from available 
assistive technology curriculums from projects and 
agencies nationwide. The curriculum is presented 
to training participants in a notebook. Additional 
resources are added as training progresses. 
Members of the state team provide training 
utilizing information from the curriculum and the 
questionnaires. In some instances, training is 
provided by outside consultants when the state 
training team cannot provide the necessary 
expertise on a specified topic (i.e. dual sensory 
impairment). 



22 



RESNA’94 • June 17.22, 1994 



PROJECT IMPACT 



RESULTS 

To date, this training model has been an effective 
vehicle for training. Critical to its evolution has 
been the recommendations and suggestions 
provided by state team members as well as 
evaluation of project activities by selected regional 
teams and the pilot project. In this way, training 
activities have been modified for more efficient 
and appropriate use of training time. Utilizing 
existing assistive technology expertise within the 
state, strengthened the networks within each 
region as well as established/identified resources 
for the regional teams within their local areas. 
The inclusion of the regional teams in the training 
activities for the subsequent regions has been 
helpful to those most recently selected. 

The initial selection process used for the regional 
team and pilot project is essential. This has 
enabled Project IMPACT to select teams or 
programs which have the necessary administrative 
support and personnel to establish and maintain 
service at project's end. In each case, the regional 
teams and pilot project have selected team leaders 
to coordinate their efforts and provide guidance to 
the team. This has been encouraged by project 
training staff but has been a natural evolution of 
team capacity building. 

The training curriculum will be revised and used 
throughout the remainder of the project. It has 
clearly demonstrated the need for an assistive 
technology curriculum which effectively addresses 
the training needs of a multidisciplinary team 
rather than individual disciplines. It was a 
challenge to use available materials to construct 
this type of curriculum. 

DISCUSSION 

As suggested by Church and Glennen (1992) the 
most essential ingredient in Project IMPACT 
training has been the identification of the 
strengths and unique characteristics of each team 
rather than trying "to do all things for all persons, 
the team should limit services to those areas which 
can be performed well". In this way, they have 
become aware of their own capacity and the 
unique capability of each regional team to their 
own community and to other regional teams. 



REFERENCES 

1. Beukelman D. & Mirenda, P., (1992). 
Augmentative and Alternative Communication. 
Baltimore, Maryland: Paul H. Brookes Publishing 
Co. 

2. Church, G. and Glennen, S: Handbook of 

Assistive Technoloev . San Diego: Singular 

Publishing Group, 1992. 

3. Cohen, C. (1986). Total habilitation and life- 
long management. In S.B. Blackstone & D. 
Ruskin (Eds.), Auementative Communication: An 
introduction. Rockville, MD: ASHA Press. 

4. Fisher P., Toszek, M., Seeger, B.E., (1993). 
Technology for Peotrle with Disabilities: A Survey 
of Needs. Assistive Technology 5.2. 106-118. 
Schoech, D., Cavalier, A., and Hoover, B., (1993). 
A Model for Integrating Technology into a Multi- 
Agencv Community Service Delivery System . 
Assistive Technology, 5.1, 11-23. 

5. Shane, H, & Yoder, D. (1981). Delivery of 
augmentative communication services: The role 
of the speech-language pathologist; Language 
Speech and Hearing Services in Schools, 12, 211- 
215. 

6. Trachtman, L. and Sauer, M. (1994). Project 
IMPACT: Development of a Statewide Service 
Delivery Model Through Local Capacity Building. 
Proposal submitted for presentation to 1994 
RESNA Conference. Nashville, Tenn. 

7. Yorkston, K., & Karlan, G. (1986). Assessment 

Procedures. In S.B. Blackstone & D. Ruskin 
(Eds.), Augmentative Communication: An 

introduction. Rockville, MD: ASHA Press. 

ACKNOWLEDGMENT 

Project IMPACT is supported in part by the North 
Carolina Council on Developmental Disabilities 
and the funds it receives through Public Law 101- 
496, "The developmental Disabilities Assistance 
and Bill of Rights Act of 1990". Project staff 
would also like to express their gratitude to Utah 
State University for the use of their materials for 
Project IMPACT training activities. 

Maggie Sauer 

Center for Development and Learn 
CB7255 BSRC 

University of North Carolina 
Chapel Hill, NC 27599 








RESNA ’94 • June 17-22, 1994 



23 



REHABILITATION ENGINEERS IN VOCATIONAL 
REHABILITATION AGENCIES: A PROFILE 



Pooraima M. Attigupam 
Center for Rehabilitation Technology Services 
South Carolina Vocational Rehabilitation 



Abstract 

This paper describes the role of rehabilitation 
engineers in vocational rehabilitation agency settings. 
Their educational backgrounds, nature of duties and 
responsibilities, and extent to which they provide 
services in specialty areas are profiled. 

Background 

Although there have been studies on job tasks and 
functions of rehabilitation service providers 
(Beardsley and Rubin, 1988), only a few have been 
conducted on rehabilitation technology professionals. 
Rehabilitation engineers differ from other engineers 
as they are required to incorporate their engineering 
skills along with their expertise in evaluating clients 
with disabilities, identifying needs and recommending 
solutions (Ellingson, 1983). However, rehabilitation 
engineers lack an identity that distinguishes them 
from others. The role confusion of rehabilitation 
engineers is not new (Trachtman, 1991). The lack of 
a conunon defuution for rehabilitation engineer’s 
qualifications and the ambiguity of their roles are 
two primary factors contributing to the confusion 
(RESNA, Jan-Feb, 1992). Because of this, people 
may tend to call themselves rehabilitation engineers 
based on their job duties rather than their educational 
background. It is not the intent of this paper to 
define a rehabilitation engineer or address 
certification and accreditation issues. Rather, the 
purpose of this paper is to provide a profile of 
rehabilitation engineers and rehabilitation technology 
professionals in VR agencies, a primary provider of 
rehabilitation technology services. This paper 
attempts to investigate the nature of the job duties, 
work experience, responsibilities and educational 
background of rehabilitation technology providers in 
VR agencies. 

Research Questions 

1. What is the educational background and experience 
of the rehabilitation engineers in VR agencies? 

2. What percentage of time is devoted to various 



activities (e.g., service delivery, 
managem^t/administrative, etc.)? 

3. How often specific types of services (e.g., 
assessment, design, training, etc.) are provided? 

Method 

A con^rehensive survey was sent to all VR agencies 
funded by the US government. The survey consisted 
of three parts: (1) an overview of the agency, (2) 
assessment/evaluation of VR clients and, (3) 
rehabilitation technology services respectively. An 
additional questionnaire accompanied the third part 
of the ^rvey with questions addressed specifically to 
rehabilitation engineers employed by or contracting 
with VR agencies. For the purpose of this survey, a 
rehabilitation engineer is defined as anyone who 
identifies him/herself as a rehabilitation engineer. 
Technology staff who considered themselves 
rehabilitation engineers were asked to complete the 
form, regardless of job title. FuU>time/part*time VR 
employees as well as outside providers returned the 
forms along with other parts of the comprehensive 
survey. For the purposes of this study, outside 
providers are considered to be those who provide 
rehabilitation technology services to a VR agency on 
a contractual basis. 

Results and Discussion 

One hundred sixteen rehabilitation engineer 
questionnaires were returned. The questionnaire 
asked respondents to report their job title. Since 
there were a large variety of job titles reported, they 
were coded into six categories. These include: 

• Rehabilitation Engineers (N = 39) 

• Rehabilitation Technology Specialists (N=36) 

• Other Engineers (N =7) 

• Admimstrator/Engineer (N=9) 

• Administrator/Other (N= 11 ) 

• Other (N= 13) 

Rehabilitation Engineers are those with a job title 
which includes the term rehabilitation engineer. 
Rehabilitation Technology Specialists are those with 
a job title related to some type of technology but not 



38 . 

RESNA ’94 • 



24 



June 17-22, 1994 



Rehab Engrs in VR 



specifically mentioning engineering (e.g.> 
rehabilitation technician, assistive technology 
specialist, etc.). Other Engineers are those with a 
reference to engineer in their job title but with no 
specific referMice to rehabilitation engineer (e.g., 
facilities engineer). Two categories of administrators 
were identified: administrator/engineers and 

administrator/other. Administrator/Engineers are 
those who have both an administrative and an 
engineering refer«ice in their job titles (lead 
engineer, senior clinical rdiabilitation engineer, etc.) 
Administrator/Other are those with an admimstrative 
referaice but no engineering reference in their job 
titles (e.g., chief, Rdiabilitation Technology). 

Others are those who could not be placed in one of 
the other categories (e.g., vocational evaluator, 
career development specialist, etc.). 

Educational Background 

Table 1 shows the highest degree earned, number of 
engineering degrees, and professional engineering 
certification held by the various rehabilitation 
technology providers. 

Table 1 : Highest Degree Earned by 

Rehabilitation Technology Providers by 

Job Title 



Job Titles for Rehabilitation Technology 
Providers 



Educational 

Background 


Rehabildation 

Engineers 


Rehab Tech 
SpedaGst 


Admin/ 

Rehab Engineer 


Other Engineer 


Admin/Other 


Other 


High School 


2 


3 


0 


0 


1 


0 


Assodate 

Degree 


0 


4 


0 


0 


0 


1 


Bachelors 

Degree 


18 


17 


5 


4 


4 


7 


Masters 

Degree 


16 


11 


2 


2 


5 


4 


Doctoral 

Degree 


3 


0 


2 


1 


0 


1 


Engineering 

Degree* 


34 


7 


10 


7 


5 


7 


Professional 

Engineers 


3 


1 


3 


2 


0 


1 


•Either Bachelors, Masters, Ph.D. separately or in combination 



From the table it can be seen that the 39 
rdiabilitation engineers hold a cumulative total of 34 
degrees in engineering. On the other hand the 36 



technology specialists hold a total of 7 degiees in 
ragineering. For all providers 10 are professional 
engineers. 

Activity Areas 

Table 2 presents a breakdown of time spent by 
rdiabilitation technology providers in various activity 
areas. 

Table 2: Average Percentage of Time Spent 

in Activitiy Areas for Rehabilitation 
Technology Providers by Job Titles 



Job Titles for Rehabilitation Technology 
Providers 



Activity Areas 


Rehabilitation 

Engineers 


Rehab Tech 
Spedafist 


Admin/Engineer 


Other Engineers 


Admin/Other 


Other 


TOTAL 


Management/ 

A^in 

AcUvilies 


10.33 


9.64 


21.89 


15.71 


42.36 


17.3 


15.2 


Service 

Ddivery 


64.33 


57.89 


33.56 


45.0 


38.18 


49.08 


54.5 


Community 

Service 


6.46 


11.89 


6.44 


14.29 


5.45 


17.77 


9.82 


Training 


9.28 


8.56 


10.22 


9.43 


9.09 


13.15 


9.56 


Research 


4.92 


6.25 


17.78 


4.14 


4.0 


6.38 


6.37 


Other 


2.61 


4.81 


10.11 


3.57 


0.91 


3.84 


3.93 



A number of analyses were performed on the data in 
Table 2. For management activities, those in the 
Administrator/Othercategory spent significantlymore 
time (M = 42.4) than did Rehabilitation Engineers 
(M = 10.3), than Other Engineers (M = 15.7), and 
than those in the Other category (M = 17.3), F 
(5,109) = 7.03, E < .001. Rehabilitation engineers 
spent significantly more time in service delivery 
activities (M = 33.5) than Admimstrators/Engineers 
(M = 64.33), F (5,109) = 3.45, p < .01. Finally, 
Administrator/Engineer spent significantly more time 
in research activities (M = 17.78) than did 
Administrator/Other (M = 4.0), than Rehabilitation 
engineers (M = 4.9), and than Rehabilitation 
Technology Specialists (M = 6.25), F (5,109) = 
2.94, E < .05. No other analyses were significant. 

Rehabilitation Technology Services Provided 

Table 3 displays a rank ordering of how often the 



39 



j<rST COPY AVAILABLE 



RESNA *94 • June 17-22, 1994 



25 



Rehab Engrs in VR 



various rehabilitation technology services are 
provided by those whose primary work activity is 
service delivery. T)rpes of services provided were 
rated on a 5 point scale according to the frequency of 
the services provided. One indicates the service was 
never provided and five indicates the service was 
almost always provided. 



Table 3: Frequency of Types 

of Services Provided 



Types of Service 


Mean Frequency 
Rating 


Evaluation/Assessment 


4.2 


Recommendaiions/Prescriptions 


4.1 


Equipment Procurement 


3.8 


Rlling/Adjustment 


3.3 


Custom Design 


2.9 


Fabricatior/Adaplailon 


2.7 


Device Training of 
Consumers/Caregivers 


2.6 


Maintenance/Repair 


2.6 


Follow-Up 


2.6 


Equipment Loan 


2.5 


Product Demonstration 


2.5 


Funding Assistance 


2.4 


Education and Training 


2.0 


Consul tail on/Technical Assistance 


1.7 


Other 


0.13 



Work Experience 

Figure 1 depicts the number of years of work 
experience as a rehabilitation engineer reported by 
those whose job title was Rehabilitation Engineer. 



Less than 




Current work focuses on the relationship between 
educational backgrounds and services provided. For 
example, does a rehabilitation engineer with a 
computer science degree tend to do more computer 
access work versus one with a mechanical 
engineering degree who tends to work with 
wheelchairs? In addition, detailed data on academic 
curricula are being collected in order to examine 
whether training programs meet the needs of 
rehabilitation technology service providers. 

References 

[1] Beardsley, M. & Rubin, S. (1988). Rehabiliiation 
service providers: An investigation of generic job 
tasks and knowledge. Rehabilitation Counsfiling 
Bulletin . 32, 122-139. 

[2] Ellingson, E. (1983). The legal responsibility of 
engineers: Where does the rehabilitation engineer 
fit in? 6th Annual Conference on Rehabilitation 
Engineering , (pp. 421-433), San Diego, CA. 

[3] Trachtman, L. (1991). Who is a rehabilitation 
engineer? RESNA 14th Annual Confei-encft (pp. 
190-192), Kansas City, MO. 

[4] RESNA. (1992). Position paper on qualifications 
and credentialing of rehabilitation engineers. 
RESNA Press . 7-9. 



Acknowledgments 

The Center for Rehabilitation Technology Services 
(CRTS) is part of the South Carolina Vocational 
Rehabilitation Department. Support for this work has 
been provided through the National Institute on 
Disability and Rehabilitation Research (NIDRR), 
U.S. Department of Education, Washington D.C. as 
part of the rehabilitation engineering research center 
grant ^133E20002-93. 

Poomima M. Attigupam 
Center for Rehabilitation Technology Services 
South Carolina Vocational Rehabilitation 
Departm^t 
1410-C Boston Avenue 
W. Columbia, SC 29171 



26 



Rgure 1. 



Rehabilitation Engineer's 
Years of Experience 



40 



RESNA ’94 • 



June 17-22 



,lS 



EST 

94 



roPY ^ • 



ASSISTIVE TECHNOLOGY USAGE OUTCOME: 
A PREUMINARY REPORT 

Frank DeRuyter, Ph.D. 

Center for Applied Rehabilitation Technology 
Rancho Los Amigos Medical Center 
Downey, CA 



ABSTRACT 

The provision of assistive technology 
appears to enhance the functional abilities 
as well as quality of life of the disabled 
individual. Service providers however 
must begin actively addressing issues of 
the appropriateness and the necessity of 
assistive technology. This interactive 
paper will report on preliminary data from 
a long term outcome study designed to 
address the issues of usage and the 
appropriateness of assistive technology as 
well as to identify the underlying and 
contributing elements associated with non- 
usage. 



BACKGROUND 

Advances in the assistive technology arena 
have moved forward a such a rapid pace that 
they have stifled many of the necessary 
changes for adequate service provision, public 
policy revisions, and the reallocation of scare 
resources. At times it appears that we are no 
longer capable of adequately keeping up. The 
increasing demands for accountability at all 
levels has become apparent. As a result, 
successful outcomes management has become 
the vogue topic of the 90's and will serve as the 
professions survival in the new century. 

Within assistive technology, very little in the 
way of outcome studies has been done. On 
the surface, the results of those studies that 
have been conducted do not necessarily reflect 
positive outcomes or successful assistive 
technology utilization. 

OBJECTIVE 

In order to adequately examine outcomes, one 
needs to look beyond traditional surface 
outcome data elements of success/failure or 
usage/non-usage of the product. Instead, there 
must be an examination of the underlying 
elements that potentially contribute to the 
ultimate outcomes. 



APPROACH 

This project developed a measurement 
protocol that enabled the capture of surface as 
well as underlying and potentially contributing 
outcome elements. Data on these elements 
has been collected at four time intervals 
(following deliveryAraining and at 6 months, 1 
year and 2 years post delivery). 

DISCUSSION 

To date, the assistive technology usage 
outcomes database includes 86 subjects who 
are at various stages in the data collection 
process. The current database encompasses a 
sex distribution of 57% male and 43% female. 
The age of the subjects at the point of initial 
data input ranges from 2 years 11 months to 
21 years 8 months with a mean age of 13 years 
7 months. The breakdown of types of 
equipment includes 59% augmentative and 
alternative communication systems, 27% 
computer systems, 9% environmental control 
units, and 5% other types of assistive 
technology. Although data collection on this 
project is ongoing, preliminary outcome data 
will be presented during this session. 

ACKNOWLEDGEMENTS 

Funding for this project has been provided ly 
the National Institute on Disability and 
Rehabilitation Research (#H133E00015). 
Opinions ejqpressed in this paper are those of 
the author and should not be construed to 
represent opinions or policies of NIDRR. 



Frank DeRuyter, Ph.D., Program Director 
Center for Applied Rehabilitation Technology 
Rancho Los Amigos Medical Center 
7601 East Imperial Highway 
Downey, CA 90242 
(310) 940-6800 

n 



RESNA’94 • June 17-22, 1994 



27 



MODIFYEVG COMVmCIALLY-AVAILABU WAlKiRS TO MEET INDIVIDUAL NEEDS 



Nffiitte S. tk^a M. Platefs and Jem Sehueh 

IMvewify of RehMlIitaden InglnegriHg Servlees 

Qiariottesville, VA 



ABsmo:^ 

This p^r discilbes two dinieal cases in whidi 
commefcially-avdlable walkers were modified to 
meet two individuals' needs. The walkers were 
ad^ted to provide the support necessary to 
in^rovetoeir posture and their ambulation. VMe 
the fflodifleafions were specific to these individuals' 
needs, they could easily be relied to others with 
similar needs. 



BACKGROUND: 

Many individuals possess mobilifympdrmentsthM 
signific^y affect fiieir daily lives. According to 
die Nadonk Center for Heddi Statistics, over 

3.500.000 persons in die United States are erdier 
unable to walk or cannot walk widiout some form 
of assistance.' ^ these peqile, ^iproxunately 

700.000 rely on walkers for mobilify. These 
individuals are of all ages and possess a variety of 
disabilides. 

Walkers are available commercially in a variety of 
sizes, styles, and configutadons, thereby fecilitadng 
a sadsfoctoty fit to most users. Unfortunately, 
commercial walkers do not meet all individuals' 
needs. For some individuals, a commercial walker 
may require custom modificadon. This was the 
case for two clients recendy seen for rehabilitadon 
engineering services. 

The first case involves a three year old girl with 
arthrogryposis. Arthrogryposis Multiplex 
Congenita is a non progressive, congenital 
ityndrome characterized by multiple contractures, 
rigid joints and sigriificant limitation of movement? 
Specifically, the giri has no functional use of her 
upper extremities, but does possess passive range 
of motion at the shoulder. She has good hip 
flexion and extension, but limited knee mobility, 
and her ankle and foot motion is normal. 

Due to her disability, this girl is unable to walk 
without assistance. Prior to Rnginppiring 
intervention, she could walk with another perspiu 



lifting her under the axillae. Wth this support, the 
giri could move her legs by fiexing her hips to 
pto^l herself forward with a gait similar to a 
soldier's march. Over time, she would become 
fotiped, and would flex at the waist, resulting in 
a poor walking ali g nment . 

The second case involves a five year old boy with 
mixed cerebral palsy that includes an atiietoid 
con^)onent. He h^ fluctuating muscle tone, 
resulting in a lack of proximal stability and an 
in^lity to sustain an t^ght posture. He used a 
Rifton a^ustable walker, v^ch includes a padded 
abdominal ring and a vinyl seat sling. The boy 
dso possesses a Guardian Strider walker which in 
its commercial form was unusable. While in the 
Rifton walker, he bore minimal weight through his 
feet, resulting in suspension at the perineum by the 
seat sling. Upon removal of the sling, he \^d 
hang by his axillae on the abdominal ring. \Wien 
the boy stood with his knees mechanically locked, 
he was able to briefly maintain proper kignment 
with minimal trunk support. Ifowever, knee and 
trunk control was inconsistent. 

OBJECTIVE: 

The objective in each case was to provide the least 
restrictive and most mobile walker for the child, 
while simultaneously inpxwing his or her walking 
posture. For the girl, die goal was to provide a 
walker that did not require upper extremity function 
for postural support. The girl's posture needed to 
be siqiported in a manner that was comfortable and 
not potentially damaging (i.e., axillary lift and 
associated crutch paralysis). For the boy, the goal 
was to incorporate a different method of providing 
vertical support to eliminate the seat sling currently 
in use on the Rifton walker. The ultimate goal for 
the boy, as written in his lEP, was to discontinue 
the use of the Rifton walker in fevor of the less 
restrictive and more practical Strider walker. 

METHOD: 

For the giri, it was necessary to provide an assistive 
device which would replace the personal attpnHant 



28 



RESNA ’94 • June 17-22, 1994 



Modifying Walkers 



ERIC 



used for ambulation. A commercially-available 
Guardian Tyke walker was chosen fijr its 
qjpropiiate size. The walker needed to be 
modified to support the giri in an i^right posture. 
The three areas needing support are the upper 
body, the arms, and the buttock. To support the 
upper body, an Otto Bock head rest was mounted 
to the fix)rrt of the walker and used as a chest 
support. The arms were supported by two 4" X 6” 
padded wooden boards placed on each side of the 
chest support, and attached with Nfiller's adjustable 
hardware. Finally, support was needed at the 
buttock to keep it fiiom drifting posteriorly. This 
was accompli^ed by adding a flat, padded gate. 
The gate swings open to allow easy entrance into 
the walker and latches shut. When the girl is 
positioned between the chest support and the 
padded gate, her body is supported in an erect 
walking posture. 

When fully supported in the walker, the child 
needs to generate the forces required to propel the 
walker. To facilitate this, the small plastic wheels 
on the fi-ont and back legs of the walker were 
removed. The fiiont wheels were replaced with 
larger mbber wheels with less rolling resistance, 
while the back wheels were replaced with mbber 
stops. Pushing the walker would have been made 
easier with wheels on the back legs, but the 
posterior force of the buttock caused the walker to 
roll backwards when these wheels were in place. 
To add incentive for pushing, pegs were added to 
the arm pads and positioned slightly in fiont of the 
hands. 

For the boy, it was necessary to provide 
^propriate siqiport that would redistribute the 
pressure exerted by the seat sling and abdominal 
ring of the Rifton walker. To accomplish this, a 
parachute body harness was used as a model. In a 
parachute harness, the body weight is supported by 
padded straps that pass fom the waist to behind 
and below the acetabula, then up between the legs 
to an attachment point on the waist. Other straps 
extend fiom the waist to pass over the shoulders, 
and then attach to the parachute itself When in 
place, the body weight is borne by the acetabula, 
with the resulting upward lifl; transferred through 
the harness to the parachute. The parachute 
harness model was redesigned for use with a 
walker by redirecting the shoulder str^s to extend 
fixsn the waist to the horizontal tubing of the 
walker, wbich provided an upward lift; and partial 
weight-bearing relief. Additional trunk support was 



provided by anterior and posterior ste^s across file 
waist. 

To fabricate this harness, tiie H^style posture belt 
by Ad^ve Engineering Labs, having a stt^ 
configuration similar to the parachute body harness, 
was modified. The modlficaticms included tiie 
addition of fisur str^s and six bucldes, reorientation 
of str^s, and the adjustment of str^ length. The 
padded shoulder str^s of the chest harness were 
used as the primary weii^-beaiing str^s tiiat pass 
below the acetabula. The harness was attached to 
the walker and fitted to the client, resulting in a 
superior walker support. 

RESULTS: 

In the case of the giri, after the walker was 
provided, she needed to be taught to move her le^ 
and to push the device forward. With practice, this 
child was able to walk across a room. Problems 
did arise with the device after use, though. Due to 
her limited strength, the girl had difficulty moving 
the walker on pluA carpet. Although the new 
wheels helped, the walker was most successful 
when used in her school, at the mall, and on other 
flat, smooth terrain. 

After a year of using the walker, this child was able 
to walk around her school, as shown in Figure 1. 
She can now maneuver it around turns, which takes 
extra effort because she is unable to lift the walker. 
As her walking and posture improved, the back 
gate of the walker was removed. Once she was 
successful at walking without the gate, the girl's 
ther^ists wanted her to be able to walk 
backwards. A vinyl band was added behind the 
buttock that could be pushed when necessary to 
scoot the walker backwards. The only other 
modifications to the walker were the removd of the 
hand pegs and the extension of the legs to 
accommodate the child's growth. 

For the boy, the harness promotes upright posture 
and redistributes the pressure to the acetabula. The 
harness interfaces with both of his walkers. When 
used in the Rifton walker, it replaces the seat sling 
yielding a much more comfortable support. When 
attached to the Strider walker, the harness provides 
the vertical support necessary for its use, as shown 
in Figure 2. With both walkers, the harness allows 
corrplete fieedom of movement during ambulation 
and significantly improves the boy’s posture and 
independent mobility. 

43 



RESNA ’94 • June 17-22, 1994 



29 



Modifying Walkers 




DISCUSSION: 

In these two cases, commercially-available walkers 
were modified to meet each child's individual 
needs. These walkers were successfully altered by 
applying forces to each child's body to obtain erect 
posture and overcome relative weaknesses. 
Currently, each child is able to ambulate on his or 
her own in a successful feshion. For the 3 year old 
gill, her walker provides for the first time the 
ability to walk on her own in a safe and 
comfortable posture. For the 5 year old boy, his 
harness allows him to attain the goal of his of 
walking with his Strider walker. 

REFERENCES 

1. Ficke, Robert C. Digest of Data on Persons 
with Disabilities . Science Management Corp. 
Greenbeh, MD. Jan 1992. 

2. Bender, Lee H and Cheryl A. Withrow. 
"Arthrogryposis Multiplex Congenita". Orthopaedic 

Figure 1 - Girl Using Modified Walker Nursing 8(5);29-35, Sept/Oct 1989. 

Nanette S. Ifiighlett 
Rehabilitation Engineering Service 
Box 35-BRH 
Charlottesville, VA 22901 
(804) 982-3845 
e-m^ njs8j@virginiaedu 



Figure 2 - Boy Using Nfodified Walker 





30 



RESNA ’94 



June 17-22, 1994 




INTEGRATION OF TECHNOLOGY INTO A NEW OCCUPATIONAL THERAPY CURRICULUM 



Aimee J. Luebben 
Occupational Therapy Program 
University of Southern Indiana 
Evansville, IN 



ABSTRACT 

While many existing occupational therapy (OT) 
schools have retrofitted their curricula with 
additional coursework in technology, other programs 
in OT have redesigned their curricula with a core of 
technology related content. This paper describes a 
developing entry-level OT program in which tech- 
nology was designed as an integral part of the 
curriculum. In this new curriculum, students spend 
a total of 365 out of 855 possible classroom hours in 
technology related content and have an additional 
150 hours of outside assignments related to 
technology. 

BACKGROUND 

Although OT practitioners have been involved in as- 
pects of technology since the inception of the pro- 
fession, entry-level OT curricula have been slow to 
add instructional \mits or additional courses in ad- 
dressing the technological needs of practicing OT 
practitioners. In a recent survey, 69% of the 
responding occupational therapists had recommend- 
ed technology during their previous two years of 
practice (1). The OT schools, however, have not 
kept current with the needs of practicing therapists. 
A survey of entry-level OT curricula found that half 
of the responding schools provide less than 20 hours 
of technology training while 25% of entry level 
curricula provide between 20 and 50 hours, and 
25% of the schools have more than 50 hours 
devoted to technology training (2). 

Entry-level OT curricula have addressed the 
addition of technology in various ways. In 1983 
Washington University was one of the first pro- 
grams to add a required OT course in technology. 
The University of Wisconsin-Madison created two 
interdisciplinary technology specialization programs, 
TechSpec in 1988 and InterACT in 1992 (3), while 
the University of Washington developed a tech- 
nology training core in the undergraduate curri- 
culum with a core course and technology units in 
other courses (4). 

In 1993 the Standing Committee of the American 
Occupational Therapy Association (AOTA) Tech- 



RESNA ’94 



nology Special Interest Section (SIS) published tech- 
technology competencies, training guidelines, and 
areas of technology content (5), This group 
developed technology competencies for occupational 
therapists at the Foundation Level, Technology 
Specialist Level 1, and Technology Specialist Level 
2. The Technology SIS recommends that OT 
curricula integrate those minimum competencies of 
the Foundation Level. 

OBJECTIVE 

To add content in technology, easting OT schools 
must either find room in courses already over- 
flowing with information or decide whether to add 
a technology class as a required or elective course. 
Either decision results in retrofitting the OT cur- 
riculum with technology content. In a developing 
OT program, however, the objective is to design 
technology, from the very beginning, as an integral 
part of the new curriculum. 

APPROACH 

Using a systems approach for integration across 
courses, the curriculum designer designated tech- 
nology as one of eight strands (health promotion 
and disease prevention, ethics, professional com- 
munication, research, cultural diversity, technology, 
collaboration with certified occupational therajy as- 
sistants, and professional conduct) to thread 
throughout the "tapestry" of the OT curriculum. 

To comprehend the inclusion of technology as one 
of eight interwoven strands, first an overview of the 
curriculum is needed. The four year OT bacca- 
laureate program is divided into prerequisite and 
professional components. After completing 59 hours 
of prerequisite courses, students apply for admission 
into the professional component wWch consists of 
65 credit hours including a minimum of 1060 hours 
of clinical experience. 

As a framework for the design of the professional 
component, the curriculum designer used Uniform 
Terminology for Occupationol Therapy (6) which 
divides human function into three occupational per- 
formance (OP) features: components, areas, and 

45 

June 17-22, 1994 31 



Teeh in a New OT Curriculum 



contexts, OP areas (OPA) include activities of daily 
living (ADD, work and productive activities, and 
play/leisure activities; OP components (OPC) 
consist of sensorimotor, cognitive, and psychosocial 
skills; and OP contexts are comprised of temporal 
and environmental aspects, Two of the five core 
courses are named after the OPAs and three after 
the QPCs, yfespan mi OP contexts are dimensions 
across the five core courses, since the performance 
of persons needing OT services varies with age and 
environment, The curriculum uses a multifrarae 
theoretical approach with five frame of reference 
categories: acquisitional, biomechanical, develop- 
mental, occupational behavior, and rehabilitation. 

Students accepted into the professional component 
must first demonstrate competency in computer 
hardware and software applications including word- 
processing, database, and spreadsheet. In integrating 
the tecimology strand into the profe^ional 
component, the curriculum designer utilized Smithls 
(7) delineation for the OT profession of two 
technology categories: support of therapy and direct 
therapeutic intervention. OT personnel use therapy 
support technologies for indirect patient care func- 
tions such as administrative, research, educatfoual, 
and documentation tasks. Smith (7) subdivides 
direct therapeutic technologies into rghaftilitativu 
and educational technologies (used fo remediate 
function) and assistive and adaptive tec^ologies 
(used to supplement or substitute for lack" of 
intrinsic function). 

For each of the 17 courses (57 credit hPUTs) which 

ppmprise the did^etie pprtipn of cwriculpm, i^ble 
I fipth the teebooiog^r codes (direct [D] and 
support [S]) and the bpurs spent in leetnre and lab 
formats as well as time spent completing outside 
assignments related tp tecimology. Students spent a 
total of 365 classroom hours (lecture and lab 
formats combined) out of 855 possible hows (17 
credit bows multiplied by the standard 15 dock 
bows pr credit how)^ Fortj^three percent of class 
time witb an extra iig bows of outside assignments 
is related to technqipp in addition to tbe didaetic 
portion of the curriculum, students spend a 
minimum of IQbO bows in glinical experiences, 
Students are ejected to integrate tbeir technology 
related information into tbeir elinicals. 

Mer receiving tbe recent publication (5) that 
included the Technology Competencies for 
Occupational Therapists, the curriculum designer 
conducted a course by course audit. Table 2 displf^s 
each technology competency for the Foundation 



Level and the corresponding OT course numbers in 
which the competency is included. In this developing 
OT program, each Foundation Level technology 
competency is addressed in two or more courses. 

IMPLICATIONS 

Accordmg to Hammel and Smith (5), the adoption 
and implementation of technology competencies 
into preservice educational program is the first step 
in the process of establishing OT within the assistive 
technology service provision arena, With the thread 
of technology interwoven throughput the preservice 
educational program, this developing entry-level OT 
program should be one of the first cwricula in the 
country to fulfil the minimum competencies in tech- 
nology, Students graduating from this program will 
have the mems to operate at the Foundation Level 
in the assistive technology service provision arena 
and the methods to seek ad^tional information to 
move to Technology Specialist Level L In a recent 
swvey (2), technology content in QT schools ranged 
from Q to 73 bows in lecture format and from 0 to 
U3 bows in lab format. With W hows of lecture, 
245 hows of lab, and 150 hows of out-side 
assignments for a total of 515 hows of tech-nology 
related information, this developing OT pro-gram 
exceeds maximum number of technology training 
hours indicated by tbe reporting OT cwricula. 

DISeU^SION 

While many OT schools have retrofitted tbeir cur- 
ricula with additional required or elective cowse- 
work in tecbnolo^ other programs in OT have re- 
designed tbeir cwricula around a core of tecbnolo^ 
related content, Cwriculum redesign and cowse- 
work retrofit are two methods that have been used 
successfully in eristing entry-level OT programs. In 
tbeir 1993-94 annual reports, all existing OT pro- 
grams must show compliance with tbe most recent 
revision of the accreditation essentials (8) which 
includes technology (essential n.B.4,g, Use of tech- 
nology in service provision and analysis of data 
when indicated). 

The developing entry-level OT program, on the 
other band, has the Inxwy of creating all 
cowsewprk from the very inception and can easily 
build tbe curriculum around technology, To improve 
technology expertise of entry-level occupationai 
therapists, the cwriculum desiper challenges other 
developing entry-level occupational therapy 
programs to plan technology as an integral part of 
tbeir curricula. 



32 



46 



RESNA ’94 * June 17-33, 1994 



Tech in a New OT Curriculum 



TfeoHNOwav eoMPurBuev 




f, 


m, 4I». 461 








|SI. 91^541, 449, 441 


hippvf 
iiil fimctfopii 


Ui,>S0, S3|, mr 
542, 440, 441 


U, stiafopw. 


M|,440 


4. TP dggljkn Pff*ir*tfo» 4*4 moipoiM^ of triapk>|gr 
«|Un or pf foctiB*. 


sip, 31?, 340, S4I, 

34?, 440. 441 


7 . tp dsfsfas 


|J|,5?P, SSI, 440. 
461.470 


g, To ptdicfM 


SSI, 44P, 460,461.470 


9. To JMyiRw (hsopnical BDdib fiom oftw firid* nfaod to 
tKfotofosr- 


SSI. 460 


tQ. Tt, itomify Mtytofl tohoM iMd* «»*«d 


SSI. 460, 461 


II: Th pnwts? 


SSI, 460, 461 


To dfOtonstiito hnuc vm of tofomtoton mmgpmmi- 
iMwokw. 


SXO. SSI, S40. 341. 
342. StO. 440. 441. 
460, 461, 470.490 


IS. 1b dAi taapnotoiT nltfid to rompntwi and compitw^- 
ipsfid ta^bswr 


S20, 330. SSI, 340, 
341. 342. 390. 44O. 
441. 460. 461,470 


14, To Iwoniiitnli huk titablidKMiting hrtmifmes. 


991,440, 470 


|i, IVt ovoltoito. mM> miMfy, uto adapt Mr fivtng tod«pl|ncy> 


lit, 991, 340. 341, 
M2. 440. 441. 461 


16. To patfoqn Ktiyiy tMhro^ featan aaahrtk, tad pmUem 
taduriqaaafbrooiaiwiito atiaoidu and waakBastaa. 


930. 991. Ml. M2. 
440. 441 


IT. 1b daaciiba bauA* aid of tedmobgy as 

ayfliid to todivilBal cHaati and aaBfop. 


331. 940. Ml. M2. 
S90, 440, 441. 460. 
461. 470. 490 


It. 1h daaciiba atokaf dilamtoai aad wiaa poacd by tectaDtogy 


940. 331. Ml. 342. 
440.441.461 


. 19. 1|. daacHbo loyoMocial imBoAim of todiBbiy wihm 


940. 331. Ml. M2. 
440.441.461 


m. 1bairirwa|ia|l7*«l>^ 


320, 461, ^ 


pg^jby tof Iwwt^s and Baitoiow of knoodidga. 


320,331,461 


7I»? 




Tf|C|>aolDgy (toa tani 'lfdtiipfc>cy* tocbdaa lahiMHaliv*. 
udaliya, aad uiuwiiwti|| tachaofagiH) Conpataacm for 
OocopattoiBl Ibecainti: Pbondatam L«w^ (Q wiSi 
cunaapoadiag OT ooum nitofiaii, 





Cpwrw NfRne (CirdKii) 


c 


LI 


u 




m 


OrimtiitHm to OT (1) 


S 


i 


0 


§ 


}}j 


ApplM (?) 


P 


10 


30 


10 


StP 


AppUe4 1 P) 


P 


s 


10 


^0 


112 


Applied l\ah W (9) 


P 


s 


10 


5 


32Q 


Pro Comnmniwdqw 


8 


9 


10 


3 




NMimdMo^Utieiiai) 


SStP 


§ 


IS 


10 


m 


Medk sod ModoUdof II 0) 




10 


m 


33 


m 


OPC It Pi^elifliKioiiil IM (S) 




3 


3 


5 


m 


UPC ?|! 8sw9fte#r swia 


9m 


li 


30 


W 


m 


epc ffli c«^ Slit 0) 


m) 


6 


10 


5 


m 


era i! ABif ws ^ 


mg 


S 


w 


10 


H} 


gm wymw 




13 


30 


10 


m 




8 


10" 


25 


10 


m 




8 


3 


3 


5 


m 


Prs II ® 


8 


3 


5 


5 


m 




9 


10 


10 


5 


•Ml 


0TR«fsrpfc@) 


§ 


9 


IS 


15 


TBTAI 




m 


241 


ia> 




ipmiiSr ownbwir m4 tee)mip$y oodes 

of V^mpyt P?=?Pireet dwapewtip 
fetenwW ^ reteN hwr* tpeot 

hi teptwpw4 feffflPti?* wed w twpp in 
m\k eiBliRmpfitit 


KEY 


e«e9BU( M^ijgruRBi IfC^kASi 
A"A88ieUMBNT 



REFERENCES 






I Somerville NJ, Wilson DJ, Shanfleld, KJ, Mack 
W. Assistive technology training needs survey, 
Assist Tseh 1990; 2; 4l»49i 

2, Kanny lA, Anson DK, Smith RO. A survey of 
tecbnolo^ education in entry’level currieuia; 
Quantity, quality, and harriers. Omp TH§f / gj/ 

Bes, im\ III 3U41R. 

3, Smith RO. InterACT; Interdisciplinary augmen- 
tative communication and tecbnolo^ tr^ning 
program. Thah SIS NmktteF W9^\ 10)5=6. 

4, Anson D. IVainiog occupational therapists in 
assistive technology! University of Washington 
program. Tech SIS Nernkmr 1993; 3(3)3-4, 

5, Hammei JM, Smith RQ. The development of 
technology competencies and training guidelines 
for occupational therapists, Am / Occup Ther, 
1993, 47(11): 970=979. 

COPY AVAILABLE BESNA ’94 * 



6, AOTA, Uniform Terminology for Occupfitionoi 
Thernpy-Third Edition (Draft IV). Rockvijle, MD: 
AOTA, 1993. 

7, Smith RO, Technological approaches to 
performance enhancement, In 6, Cbristiapsen 
and C, lawn, eds, UmpsHeml thempy: gyer- 
gming Hamnn perfgrmmf^e defioits. Thorofare, 
NJ; SLACK, 1991: 747=786. 

I, aOTA. issentials and guidelines for an accredi- 
ted educadonal program for the occupational 
therapist, Am / Ocmp fher, l^L 45(12): 1077= 
1084. 

Aimee J, Luebben, Director 
Occupational Therapy Program 
University of Southern Indlaua 
8600 University Boulevard 
EvansviUe, IN 47712, 812=464-8600 

47 

me 17-22, X994 



33 



HEART . A STUDY ON ASSISTIVE TECHNOLOGY IN EUROPE 



Gunnar Fagerberg and Tomas Lagerwall 
The Swedish Handicap Institute 
VaUingby, Sweden 



ABSTRACT 

HEART is a study of different aspects on assistive 
technology in Europe, under the TIDE 
(Technology Initiative for Disabled and Elderly 
People) programme in Directorate-General XIII in 
the Commission of the European Communities. 
The work is carried out by a Consortium of 21 
members from 12 countries under the leadership of 
The Swedish Handicap Institute as Main 
Conttactor. 

At the time of writing, the study has completed its 
first six months, out of eighteen in total. The first 
period has been devoted mainly to administrative 
preparation, planning, methodological 
development and data collection. By the time of 
presentation, results will have been produced in 32 
reports. 



BACKGROUND 

HEART (Horizontal European Activities in 
Rehabilitation Technology) is a study of different 
aspects on assistive technology in Europe, under 
the TIDE (Technology Initiative for Disabled and 
Elderly People) programme in Directorate-General 
XIII in the Commission of the European 
Communities. 

The HEART study is carried out by a consortium 
made up of 21 institutions, organizations and 
companies under the leadership of The Swedish 
Han^cap Institute as Main Contractor. 

The overall objectives of the study are: 

- to survey, analyze and assess the current 
situation 

to spread information about the current 
situation 

• to create communication channels between 
actors in Europe 

to show routes to facilitate the creation of a 
single market by proposing directions and 
priorities. 

The ultimate goal of the study is to improve the 
life situation for persons with disabilities and 
elderly persons. Products should have a high 



quality and utilize current technology. They should 
be accessible and affordable. Users must have 
purchase power and influence at different levels. 
There should be a general accessibility for persons 
with disabilities to products and services offered in 
society. 

The study covers six areas, or lines. Each line has 
a designated Line Leader and line members. The 
Line Leaders and the Main ConU’actor form the 
Management Committee for the study. The study 
started on April 1, 1993 and will finish by 
October, 1994, 

METHODS 

The first six months of the study have been 
dominated by organization, planning, preparations 
and preliminary collection of data and facts, for 
which a number of methods have been used. The 
consortium has established a common terminology 
and definitions. Data collection procedures have 
been coordinated. Interdependencies and areas of 
common interest between two or more lines have 
been identified and ^reached. 

In order to attain a higher coherence in the study, 
some common areas to be studied in all lines have 
been selected. They are: user involvement and 
influence, wheelchails and new media. 

A Consumer Board has been established with the 
purpose of facilitating a direct dialogue between 
the HEART consortium and the users of assistive 
technology. It is comprised of one representative 
of each of six mqjor organizations of disabled 
people. 

RESULTS 

The results reported in the first round of 
deliverables contain a wealth of information about 
rehabilitation technology in Europe. Much of it is 
new information, information never compiled 
before or information previously not easily 
accessible. 

Line A: Standards, testing and certification/ 

specification, ol rebabiiitatiQOLtechaolflgy 

48 



RESNA’94 • June 17-22, 1994 



34 



HEART - a Study on Assistive Technology in Europe 



A survey of laboratories involved in testing 
assistive technology in Europe has been carried 
out. Forty-one laboratories in 13 countries 
reported that they are involved in testing 
assistive technologies. The largest number of 
laboratories identified is found in Germany, with 
six laboratories, and the United Kingdom, with 
five laboratories. The geographic spread is rather 
wide, but no laboratories have been identified in 
Luxembourg, Portugal, Spain or Greece. 

Existing standards and current standardization work 
in assistive technology were surveyed. The work 
resulted in a complete list of all existing national, 
European and international standards and current 
standardization work in assistive technology. At 
the international level (ISO and DEC), around 30 
standards exist and another 50 are drafted and 
planned in the work programmes. Most of them 
concern technical aids for personal mobility and 
orthoses and prostheses, 17 standards or draft 
standards deal with orthoses and prostheses, 23 
with wheelchairs. European (CEN and 
CENELEC) standardization work in the assistive 
technology field covers at present nearly 40 draft 
standards, or standards planned in the work 
programmes. The national survey shows a very 
mixed picture. Most of the standards are equal to or 
based on international standards. 

Information technology and telecommunications: 
At the international level, no existing standards 
were found. Only three drafts and several Work 
Items were found. Neither were any existing 
standards found at the European level. At the 
national level some existing standards were found 
which concern Braille Graphic Characters and 
orientation aids. 

A general model for user influence in 
standardization work has been developed. A field 
trial with user participation in the standardization 
process has started. The group of products to be 
standardized is telephone keypads. 

Line B: Coherence between and amgng 
rehabilitation technology industrial sectors 

Information on links, networks and other contacts 
among assistive technology industries, mainstream 
IT&T industries, and non-industrial organizations 
has been gathered through literature searches, 
interviews and an extensive questionnaire. 

A remarkable amount of cooperation was found in 
specific countries and regions (Northern and central 
part of Europe) but much less in other countries 
and at a pan-European level. Networking, -in 



particular, seems to be a powerful tool to improve 
the situation of smaller companies. The direct 
involvement of end users is generally weak but 
some significant exceptions were found. There are 
numerous links between mainstream IT&T 
industries and assistive technology companies in 
research and development activities. 

Line C: Rehabilitation Technology Servic_e 
Delivery 

Relevant aspects and characteristics of service 
delivery systems were identified. A search for 
literature and other information about the service 
delivery systems in the 16 countries under study 
was made. 

For each country, there will be a general 
description of the assistive technology service 
delivery system or systems. In addition, the 
procedure for the provision of five selected product 
categories will be described in detail: manual 
wheelchairs, telephones for the deaf, computer 
peripherals for the blind, hygiene aids, and 
augmentative communication devices. 

Draft descriptions of the systems in all the 
countries under study were made. National 
contacts in each country completeted and corrected 
the information. One or more national 
organizations of the disabled studied and 
commented on the national descriptions. The 
result is a compilation of detailed descriptions of 
assistive technology provision systems in a large 
number of European countries. 

Line D: Legal and 

Economic Factors Impacting Rehabilitation 
Technology Availability 

A census of legislation affecting the availability of 
assistive technology products has been made. 
Legislation was studied in 11 countries in the 
following areas: Daily living. Education, Working 
life and vocational training. Social protection. 
Transport, and Housing. 

There are considerable differences in the policies 
adopted towards legislation for the elderly and 
disabled in the countries studied. However, there 
are also a number of common themes. In terms of 
specific legislation, three broad models have been 
identified. Different philosophies behind the 
legislative approaches have been found, as well as 
a distinction between the legislative bases for the 
provision of medical devices and the provision of 
assistive devices. 

49 



RESNA ’94 



June 17-22, 1994 



35 



HEART - a Study on Assistive Technology in Europe 



A theoretical study of microeconomic concepts and 
a survey of existing socio-economic models has 
been performed. Three kinds of microeconomic 
models were identified: Cost-Benefit, Cost- 
Efficiency, and Cost-Utility. The main advantage 
of these models is that they permit social aspects 
to be integrated in economic models. However, the 
application of these models is limited. In fact, they 
are specific to a product or to a target group or to 
an economic actor. 

Only a few studies were found in the survey and 
the ways of describing outcomes differ. 
Comparisons seem to be very difficult to 
accomplish, unless standardization is at hand, but 
there is a long way to achieve this. Cost- 
effectiveness comparisons based on quality of life 
cannot be made to^y. There is one unique way out 
of the problem, the use of cost-utility analysis. 
Some studies using cost-utility analysis are 
available in the field of assistive technology. 

Line E: Rehabilitation Technology Training 

A survey of existing training programmes for 
engineers/technicians and related professions 
(therapists, physicians, teachers, etc, with an 
emphasis on occupational therapists) in Europe 
and North America has been cairi^ out. 

Pre-service training in assistive technology of 
technicians and engineers in Europe was mainly 
identified in two European countries: United 
Kingdom and Sweden. In Italy, The Netherlands, 
Portugal, Spain and Denmark, some components 
of assistive technology training are incorporated in 
courses at a university level. 

In North America, several universities with pre- 
service training in assistive technology for 
technicians and engineers, mainly at a Master "s 
level, were identified. 

Line F: Emerging Areas of Potential 

Rehabilitation Technology Research and 
Development 

The following actions have been taken and 
reported: 

Selection of functions to be performed for social 
integration and technologies to be considered in the 
study. 

Selection of application scenarios - environments 
which will change significantly in the near future, 
due to changes in technology or other causes. 
Assessment of the present situation in assistive 
technology: what support is currently available to 



people? 

Identification of the problems of people with 
disabilities in performing the selected functions 
with the technical support presently offered. 
Identification of developments in assistive 
technology and in technology not used in 
£^plications for persons with disabilities, but with 
a potential impact. 

Identification of possible impact of these 
developments on the situation of people with 
disabilities and elderly people. 

Identification of critical issues that could influence 
technology transfer in assistive technology, 

REFERENCES 

1, Fagerberg, G, and Lagerwall, T, Main 
Contractor First Progress Report, HEART Study 
(TIDE 309), The Swedish Handicap Institute, 
Vailingby 1994. 



Gunnar Fageiberg 

The Swedish Handicap Institute 

Box 510 

S-162 15 vailingby 
Sweden 



50 



RESNA’94 • June 17-22, 1994 



36 



CONSIDERING QUALITY IN A CROSS-DISCIPLINE 
REHABILITATION ENGINEERING SERVICE 

Gilbert D. Logan David F. Radcliffe 

Royal Brisbane Hospital Mechanical Engineering Dept 

Queensland, Australia University of Queensland, Australia 



Abstract 

In response to demands/expectations of clients 
providers of rehabilitation engineering services 
must become quality conscious and apply the same 
quality— generating tools that are becoming 
commonplace in manufacturing industries: Quality 
Function Deployment, Concurrent Engineering, 
Continuous Process Improvement, to name a few. 
Quality techniques are being investigated in a 
rehabilitation engineering service that manufactures 
customized mobility and seating. The aim of the 
research is to: (1) ascertain how these techniques 
can be best applied in a cross-discipline 
rehabilitation engineering team and, (2) develop 
measures of outcome for the client which 
reference the inputs to the rehabilitation 

engineering process as well as the quality of 
service and products supplied to the client. Initial 
results show Quality Function Deployment and 
Concurrent Engineering offer considerable benefit 
to a cross-discipline team endeavouring to 
increase service and product quality. 

Background 

Quality is a catchword of the 90‘s. Consumers 
demand quality in goods and services; but what is 
quality? How should it be measured? Garvin ^ has 
considered five approaches to defining quality: the 
Transcendental approach of philosophy. Product- 
based approach of economics. User-based 
approach of economics, marketing and operations 
management. Manufacturer-based approach. 
Value-based approach of operations management. 
These approaches have relevance to rehabilitation 
engineering service providers whose clients have 
rights to receive quality services and products. 
Purchasers of rehabilitation engineering services 
are increasingly discerning and apply the same 
purchasing decisions they apply to other consumer 
products. Rehabilitation engineering service 
providers who manufacture customized devices 
(e.g seating and mobility devices) for clients, 
particularly severely disabled clients, face 
difficulty in ensuring the outcome for the client is 
optimum and matches goals that have been agreed 
to by client (or carer) and the rehabilitation 
engineering team. Deviation from prescription or 
design specification decreases the quality of 



outcome for the client, and portends loss of 
revenue and loss of future business for the service 
provider. 

Objective 

The objective of this study is to determine what 
quality techniques can rehabilitation engineering 
services employ that will increase the likelihood 
of success in the provision of a technological 
solution to a client's physical problem or 
difficulty? 

Approach 

A rehabilitation engineering team has been 
investigating some techniques used in 
contemporary engineering practise to evaluate 
their usefulness in rehabilitation engineering. 

Flowcharting 

Flowcharting involves mapping the inputs, 
inventories, outputs and activities that occur in 
rehabilitation engineering practise and linking the 
various components to show flow of the process 
and relationships. From a flowchart can be 
gleaned (1) a critical path of elements or 
activities that must be in place in order to avoid 
delays, (2) labour and time requirements for 
activities, (3) the inter-relationships of the 
process, (4) possibilities for improvements in 
service delivery, (5) cost structures. 

Quality Function Deployment (QFD) 

QFD is a methodology for selecting and 
prioritizing quality requirements of the customer, 
converting requirements into technical/design 
criteria and systematically utilizing these criteria 
to drive the product/service process to achieve the 
customers' requirements QFD is a powerful 
technique for identifying and integrating 
disabled clients' requirements for quality into the 
processes and activities of the rehabilitation 
engineering effort. 

Concurrent Engineering (CE) 

Concurrent Engineering describes the concept of 
integrated, simultaneous design of a product and 
its related processes, e.g. manufacture, testing."^ 
CE embodies a team approach to problem solving. 



51 

RESNA ’94 • June 17-22, 1994 



37 



Quality in Rehabilitation Engineering 



The combination of QFD and CE is a powerful 
tool for rapidly producing customized devices with 
quality and features meeting the requirements of 
the disabled client. 

These techniques are being utilized in a seating 
clinic which has operated for many years with 
limited resources. This clinic manufactures 
customized mobility and seating devices for adults 
and children with severe disabilities. A large 
backlog of work threatened to overwhelm the 
facility and staff, prompting an analysis of the 
modus operandi of the team. 

Analysis of video recordings of client assessment 
sessions (which operated on a medical model at 
the time) revealed this method of operation was 
insufficient to acquire the extent of information 
necessary for understanding the client's 
requirements^. 

Flowcharting revealed the clinic was not making 
best use of labour resources by seeing clients for 
short fitting sessions between bursts of 
manufacturing effort. Members of the cross- 
discipline team (rehabilitation engineer, orthotist, 
physiotherapist, occupational therapist, technician) 
were frequently absent from important processes; 
assessment, prescription decisions and trial fitting 
of manufactured devices. Important skills and 
information input were missing from processes 
which had a direct bearing on design, manufacture, 
outcomes and overall quality. 

Equipment manufactured by the clinic was 
frequently not sufficiently robust to withstand the 
rigours of use, particularly in large residential care 
institutions. Designs, materials of manufacture did 
not reflect client and carer needs in equipment 
such as tilt-in-space, lightweight, compact, easy 
to use. 

Results 

Major changes were made to the clinic reflecting 
an emphasis on the outcome for the client. This 
list is interim as research into quality aspects of 
service delivery is on-going: 

1. Medical style assessments were replaced with a 
format which aimed to acquire pertinent, accurate 
information and data about the client and their 
environment in which prescribed equipment would 
be used. The imporUmce of the team has been 
engendered in assessment technique. 

2. Client evaluation, prescribing, decision making. 



design, manufacture and trialing were 
programmed to occur on one day, occupying the 
whole day. Flowchart and resource analysis 
indicated 2 clients requiring this level of input 
could be completed per week. 

3. Documentation was developed to match and 
prompt the progress of the assessment process for 
detailed, accurate information and data gathering. 

4. Goals for the equipment and the client's use of 
it were established by the team, elucidated to 
client (or carer) and used comparatively during 
the operation. 

5. QFD techniques^ were employed (and continue 
to be) to sort information from disabled clients 
about equipment features that need to be designed 
in and how this should be done. 

6. CE practice was introduced to speed up 
production. Where possible design and 
manufacture are performed as parallel activities 
instead of serial activities. The program aim is for 
the client to leave the clinic at the end of the day 
with a device to trial at home for a week, to 
isolate defects and operating difficulties before 
completing manufacture (e.g. upholstery). 

Discussion 

Application of contemporary engineering process 
and quality techniques, facilitated by willingness 
of team members to reflect on the results of 
analysis and make changes has significantly 
improved productivity of the clinic. A service that 
had a backlog of work approaching 12 months is 
now able to clear 200 clients per annum, 
manufacturing 130 major and 55 minor devices 
(1993 output). 

Video analysis of client assessments, team 
discussions and decision-making is particularly 
useful for revealing loss of information that is 
relevant to the client's situation and requirements. 
Information capture, the assimilation and 
comprehension of this information and its 
manipulation or conversion into technical data is 
crucial to the success of downstream activities 
such as rehabilitation device design and 
manufacture. Good documentation of client 
information, assessment findings, team discussion 
and decisions in a format that is readily 
discernible by professionals from physical 
sciences and life sciences helps accuracy and 
reduces the likelihood of misinterpretation. 

52 

June 17-22, 1994 



38 



RESNA ’94 



Quality in Rehabilitation Engineering 



Some client data is best kept visual as it involves 
mimicry or similar highly visual detail. A 
photograph does not capture the motion of this 
detail and attempting written description loses both 
content and context. The retention of video data 
(video clips) in a client’s file on a computer 
database is being examined. A major detraction is 
the time required to edit the video to find pertinent 
frames of video plus appropriate memory to store 
video data for rapid retrieval by computer 
(writeable CD ROM has application). 

QFD’s power in rehabilitation engineering is in: 

(a) converting client quality and design features 
into engineering considerations with analytical 
connotations for inclusion in product design and, 

(b) providing a mechanism for assisting the 
inclusion of client— inspired quality features at each 
stage of the rehabilitation engineering process. 
This could be a vital tool in rehabilitation 
engineering because able-bodied designers often 
do not recognize subtleties unique to disabled 
persons which must be accommodated in devices. 

Rehabilitation engineering is a fertile area for 
unique application of quality techniques. The team 
approach has been well established in 
rehabilitation engineering, particularly where a 
service has been founded on a medical model. 
QFD and CE being team-centred techniques lend 
themselves to being utilized in rehabilitation 
engineering. It must be recognized that QFD, CE 
and other quality techniques were developed for 
large volume manufacturing industries (consumer 
products such as automobiles, refrigerators). 
Modification of these engineering quality 

techniques is necessary to accommodate both 
physical science and life science nature of the 
disciplines associated with rehabilitation 

engineering teams and the one-off nature of 
device development and manufacture. 

Conclusion 

Our work so far has found that quality techniques 
applied to a cross-discipline team rehabilitation 
engineering service work favourably and help the 
team achieve a level of quality in client services 
that a conventional medical model team has been 
unable to achieve. Tools such as QFD and CE 
developed for manufacturing industries require 
adaptation for application to the type of group 
found in rehabilitation engineering. 



Future Work 

An in-depth study has commenced of the 
moment-to-moment activity, communication and 
process during client assessment, team decision- 
making and design/manufacturing tasks in a 
seating clinic, using video, observation and 
examination of case notes. 



References 

1. David A. Garvin, What Does ’’Product Quality” 
Really Mean?, Sloan Management Review, Fall 
1984, pp25-43. 

2. Y. Akao, Quality Function Deployment, 
Productivity Press, Cambridge, MA, 1991. 

3. Gary S. Wasserman, Integrated System Quality 
Through QFD, Institute of Industrial Engineers, 
1989 HE Integrated Systems Conference & 
Society for Integrated Manufacturing Conference 
Proceedings, pp 229-234. 

4. }, Turino, Managing Concurrent Engineering: 
Buying Time to Market j a Definitive Guide to 
Improve Competitiveness in Electronics Design 
and Manufacturing, Van Nostrand Reinhold, New 
York, 1992. 

5. D.F. Radcliffe, P. Slattery, Emergence, learning 
and Inter-reaction in a Cross-Discipline Design 
Environment, ASME Design Technical 
Conference, 1992. 

6. Ray Thackeray, George Van Treeck, y^plying 
QFD for Software Product Development, Journal 
of Engineering Design, Vol 1, No 4, 1990, pp 
389-410. 

The assistance of Royal Brisbane Hospital and 
Royal Children’s Hospital in providing facilities 
for this work is acknowledged. 



Gilbert Logan 

Bioengineer-in-Charge 

Royal Brisbane Hospital 

RBH Post Office, Queensland 4029 

Australia. 

phone +61 7 2537412 
fax +61 7 2531389 

e-mail: logan@sun.mech.uq.oz.au 



RESNA ’94 • June 17-22, 1994 



39 



rehabilitation engineering training - THE PRACTICAL SIDE 



Blair A Rowley 

Department of Biomedical and Human Factors Engineering 
Wright State University 
Dayton, Ohio 45435 



ABSTRACT 

This paper presents the practical side of a one 
year masters level training program. It addresses 
how a diverse group of students are being 
prepared to provide applied engineering s kills . 
Course contents and how practical aspects are 
employed are discussed. The results of an 
internship utilizing a Rehabilitation Engineering 
Service are also presented. 

BACKGROUND 

Rehabilitation Engineering educators in the 
United States are in the process of defining 
educational goals, learning experiences, and 
performance expectations. At the 1993 RESNA 
Annual Conference, a diverse group of 
engineering educators and interested parties met 
to share concerns and ideas. They decided that 
a workshop on Rehabilitation Engineering 
Education should be held to pull together the 
growing Rehabilitation Engineering education 
community, produce stronger educational 
programs, and identify the educational 
requirements needed to attract, retain, prepare, 
and graduate Rehabilitation Engineers to meet 
the increasing challenge of applying complex 
technology to the needs of individuals with 
disabilities. Some of the training issues may 
focus on knowledge in life science, human 
factors/ergonomics, applied engineering, 
rehabilitation service deliveiy systems, and an 
internship experience. This paper discusses the 
approach and experiences in teaching the 
practical side to students in the masters level 
Rehabilitation Engineering Training Program at 
Wright State University (WSU). Hopefully this 
will be of assistance to other students, educators, 
and the workshop participants when it is held. 

In a previous paper, the four quarter, one year 
program in Rehabilitation Engineering Training 
at WSU was described (Rowley). Over the past 
two years eleven students have graduated and 
this year six more will complete the program. 
There are also several part time students. Three 
graduates are working at rehabilitation centers, 
one started a rehabilitation engineering company, 
one went to medical school, one is completing a 



militaiy obligation, one is the ergonomist for a 
division of an automobile company, two are 
consultants, one works for a prosthetics firm, and 
one is working in a family business. They 
entered the program as engineers with 
biomedical, human factors, electrical, mechanical, 
and engineering physics backgrounds. Three 
came from industry, two from other graduate 
programs and the others immediately from their 
BS programs. The full time students the first 
year were four men and one woman; the second 
year, four men and two women; and the third 
year, one man and five women. 

METHOD 

The Practical Side 

To meet the challenge presented by this diverse 
group of students, analytical and applied courses 
have been developed so that all will have a 
similar set of skills and experiences upon 
graduation. The students are expected to enter 
the program having already taken courses in 
anatomy and physiology. The analytical courses 
are: 

Advanced Biomechanics 

Applied Statistics 

Bioinstrumentation 

Biomechanics 

Biomedical Computers 

Neuromuscular Rehabilitation 

Robotics 

The applied courses are: 

Human Factors Engineering in Rehabilitation 
Introduction to Clinical Practice 
Introduction to Rehabilitation Engineering 
Introduction to Rehabilitation Design 
Rehabilitation Assistive Systems 
Rehabilitation Engineering (the internship) 
Rehabilitation Egr Computers 
Rehabilitation Egr Design 
Rehabilitation Egr Service Deliveiy 

These focus on communication skills, social skills, 
team skills, applied computers, applied 
electronics, applied mechanics, and applied 



RESNA ’94 



54 

June 17-22, 1994 



40 



REHAB. ENOINEERINO TEIAININO 



ergonomics. Within these courses the students 
learn about assistive technology; wheelchairs, 
seating, driving, augmentative communications, 
and computer adaptations. They learn how to 
apply basic engineering building blocks such as 
CMOS devices and four bar linkages; the range 
of disabling conditions: CP, MS, SCI, 

amputation, head injury, aphasia, etc.; and how 
rehabilitation delivery service works. By the 
thir d quarter they are receiving applied 
instruction on home and workplace 
modifications, meeting the needs of rehabilitation 
counselors and their clients, and receiving an 
introduction to a rehabilitation center; they will 
have evaluated, designed, and constructed 
electronic and mechanical devices, produced at 
least three major reports and made numerous 
presentations; and as teams they will have 
designed and delivered several solutions to 
problems for individuals with disabilities. 

Through out the year field trips help bring to 
focus class room instruction. These trips involve 
vehicular modifications, an industrial setting for 
the blind, a workmens’ compensation 
rehabilitation center, schools and assembly lines 
at a county MR/DD board, prosthetics and 
orthotics manufacturing, a rehabilitation center, 
a center for augmentative communications and 
computer adaptation, a truck plant with an aging 
work force, the Ohio Rehabilitation Technology 
Association meetings where case studies are 
presented, and the RESNA annual conference. 

Guest speakers are invited to present in the 
Introduction to Rehabilitation Engineer, the 
Rehabilitation Engineering Systems, and the 
Introduction to Clinical Practice courses. These 
speakers come fi-om the medical profession, 
vocational rehabilitation, services for the blind, 
services for the deaf, county boards for MR/DD, 
workmens’ compensation rehabilitation division, 
and practitioners within rehabilitation service 
delivery. 

Timtmcttofial Resources 

The following texts have been used with the 
applied courses. 

"Americans with Disabilities Act Handbook", 
Fqiial Employment Opportunity Commission and 
the U.S. Department of Justice 

"Augmentative Communications an Introduction", 
American Speech-Language-Hearing Association 



"CMOS Cookbook", 2nd ed., by Lancaster, 
Revised by Berlin, SAMS 

"Cumulative Trauma Disorders: A Manual for 
Musculoskeletal Diseases of the Upper Limbs", 
by Bristol, Tavlor & Francis. Inc, 

"Electronic Devices for Rehabilitation", by 
Webster, et. aL, Wilev Medical 

"Fundamentals of Industrial Ergonomics" by 
Paulat and Mustafa, Prentice-Hall 

"Mechanism Design Analysis and Synthesis 
Volume I" by Erdman and Sandor, Prentice-Hall 

"Positioning for Function", by Bergen, Presperin 
and Tallman, Valhalla Rehabilitation 
Publications. Ltd. 

"Using Toolbook" and "Using OpenScript", 
Asymetrix 

The students also receive course handouts which 
have been developed by the instructors. Subjects 
include disabilities, medical aspects, service 
delivery, vehicular modifications, assessments, 
wheelchairs, the ADA, the Rehabilitation Act 
with Amendments, and the Technology Act 
There is also an adapted computer laboratory on 
campus with the latest technology installed and 
working. In the Rehabilitation Engineering 
Laboratory (REL) are examples of augmentative 
communications devices. 

Each student is provided a large filing drawer 
within the REL and encouraged to start building 
their own Ubrary using the REL resources as a 
guide. These resources include pamphlets fi’om 
manufactures, catalogs, proceedings, journals, and 
books on rehabilitation and engineering. A few 
of these are Assistive Technology, Augmentative 
and Alternative Communication, Journal of 
Rehabilitation Research and Development, IEEE 
Transaction on Rehabilitation Engineering, 
Technology and Disability, Technology Aid to the 
Disabled Journal, Oosing the Gap, Solutions- 
access technologies for persons who are blind, 
Technology for Independent Living Sourcebook, 
Recreation-Oriented Assistive Device Sources, 
Choosing the Best Wheelchair Cushion, Design 
Engineering Handbooks, and Hyper-AbleData. 

The Tmarnahto 

This capstone course as previously described has 
been restructured to give the students an 



r 55 

RESNA ’94 • June 17-22, 1994 



41 



REHA& ENOINEERINO TRAINING 



improved experience providing rehabilitation 
engineering service (Rowley). The clinical task 
list is still required and has been improved upon 
with the help of the rehabilitation center staff. 
The students are still graded upon attendance, 
completion of the clinical tasks, and their 
designs. The major change is how the internship 
is organized and managed. Each day is now 
divided into two parts. The mornings are 
devoted to the clinical tasks, rounds, and 
rehabilitation teaming. In the afternoons a 
Rehabilitation Engineering Service (RES) is 
operated. The RES responds to the needs of 
the rehabilitation center staff: inpatient and 
outpatient services; home health care, and 
industrial evaluations; computer, wheelchair and 
driving adaptations; and equipment modifications. 

This past summer 51 projects were considered 
by the RES. Forty-three of these entered the 
preliminary design stage, 33 were designed, and 
17 were completed and delivered. Some of 
these were an automobile simulator for transfer 
training, adaptive equipment setup for a person 
with SCI, a set of adaptive switches, modification 
to a fluoroscopy chair to aid swallowing 
evaluations, a new fixture for the Baltimore Test 
Equipment Work Simulator to provide therapy 
for vertical motions, a simulator to train patients 
to use the Life Line, and a putting green. 

In addition, each student completes at least two 
home visits, one workplace evaluation, and one 
service call for vocational rehabilitation. The 
student prepares the reports that are given to 
the counselors. They also participate in 
presenting a workshop for rehabilitation 
counselors. This serves as practical application 
for their communication skills and helps increase 
their networking base. 

DISCUSSION 



with a set of experiences that prepare them to 
apply technology to the needs of rehabilitation 
service agencies, schools, industry, providers, and 
individuals with disibility. 

REFERENCES 

Rowley, B.A., "The Rehabilitation Engineering 
Training Program at Wright State University", 
Proc. RESNA ’93 Annual Conf.. pp. 54-56, 1993. 

ACKNOWLEDGMENTS 

The contents of this program were developed 
under a grant from the Department of 
Education. However, those contents do not 
necessarily represent the policy of the 
Department of education, and you should not 
assume endorsement by the Federal 
Government 

Many thanks to the service providers at the St 
Elizabeth Rehabilitation Center in Dayton, Ohio, 
the guests speakers, and the field trip hosts that 
help make this program so practical. 



Blair A. Rowley, Ph.D., P.E., C.C.E. 

Biomedical and Human Factors Department 

Wright State University 

Dayton, Ohio 45453 

(513) 873-5073 Office 

(513) 873-5009 FAX 

e-mail: browley@matrix.wrightedu 



The practical side of this training program 
develops the applied skills of the Rehabilitation 
Engineer. Communications are honed through 
reports and presentations. Engineering skills are 
developed through courses with real life 
application requirements; students learn by 
doing. Networking, so important for developing 
a service practice, is begun through field trips 
and guest lectures. Students are exposed to a 
large number of individuals with disabilities and 
are required to become very familiar with the 
service delivery system," how a rehabilitation 
center functions, and the delivery of service in 
the home and workplace setting. They graduate 



56 



42 



' c 

RESNA ’94 • June 17-22, 1994 



IN-SERVICE AND CmSUMER TRAINING IN ASSISTIVE TECHNOLOGY? 
A WORK IN PKX3RESS 

Nancy Friedrmn & Mary Brady 

MD TECHNOLOGY ASSISTAJCE PROGRAM 
Baltimore, MD 



ABSTRACT 

Due to the interdisciplinary nature 
of Assistive Technology training, 
there are few professional pre-service 
training programs in colleges and 
universities. Maryland *s Technology 
Assistance Program’s response uses a 
multifaceted training-consultant 
model* Training is offered in each 
region of the state and covers topics 
including sensory aids, cutting edge 
technologies, funding, recreation 
and leisure, environnental modifica- 
tions and technology for kids. 
Evaluations and needs assessment are 
done of each training program to 
insure quality of the training and 
continued iirprovonent , or development 
of new programs. 



THE PROBLEM 



degrees. Courses were found as parts of 
professional training programs. There 
are few exceptions to the above. Johns 
Hopkins University offers a degree in 
assistive technology for educators. The 
University of WisconsirvMadison offers 
several degrees in rehabilitation-related 
fields with training i^ assistive tech- 
nology. The University of Pittsburg is 
planning to institute a new department 
called Rehabilitation Science and Tech- 
nology. Presently, a certificate is 
offered to Occupational Therapists (OT), 
Physical Therapists (PT), Engineers, and 
Biomedical Engineers. (Brienza 1994). 

A recent literature search found very 
little published information about vhat 
kind of training is being done at the 
professional level in assistive technology. 

One might ask, ”Why the focus on 
professionals?” Peterson, MacArthur & 
Brady (1991) (3), researchers in Maryland 
found that most consumers of 
AT learned about their devices from 
medical service providers. 



Assistive Technology training enconpasses 
the skills and knowledge of many different 
disciplines: including occupational 

therapists (OT), physical therapists (PT), 
speech therapists, engineers, teachers, 
nurses, physicians, rehabilitation 
counselors and computer experts. Even 
so, there are surprisingly few assistive 
technology (AT) concentrations in pro- 
fessional pre-service training programs. 
Courses are offered in programs for 
occupational therapists, nurses, engineers, 
physical therapists and speech therapists. 
Some special education programs have a 
course in AT. There are bio-medical 
engineering programs but entire programs 
for rehabilitation technologists are few. 

In reviewing lists developed by Training 
Programs in AT (1992) (1) and Technology 
Training: Resources for the Trainer (1990) 

(2), no programs were listed as offering * -4 

• if* 



RESNA ’94 



NEW APPROACH 

Maryland’s Technology Assistance Program 
has responded to the needs for training 
of professionals by using a multifaceted 
consultant model. Generalists with exper- 
tise in a focal area develop two hour 
training modules. Each module is offered 
several times a year in different regions 
of the state. In this way all training 
programs are avilable within a reasonable 
commute to all citizens in the state. 

Consultants available through the 
business comnunity, and organizations 
such as RESNA, are utilized vhere needed. 
Outside their focal area of expertise, 
staff mOTbers are encouraged to attend 
these training programs, along with 
members of the comnunity. In this way, a 
broadly-based consultant network is 
reinforced, as all staff menbers acquire 
knowledge of additional areas of technology. 

57 

June 17-22, 1994 



43 



Training in Assistive Technology 



Technology seminars cover topics of 
sensory aids, augfnentatlve cortmunlcation 
devices, funding for assistive technology, 
technologies for adaptive recreation and 
leisure, technology for kids, environmental 
modifications, and "cutting edge" 
technologies. 

Educational approaches are designed 
for the adult learner and many programs 
use a hands-on, as well as a lecture format. 

Each program is evaluated on content 
and presentation. Data is kept on 
attendance and consumer satisfaction. 

Part of the evaluation asks the audience 
vJiat they would like in future training 
programs. In addition, needs assessments 
are conducted in conjunction with 
state agencies, and both private for profit 
and private not for profit rehabilitation 
facilities, persons viho deliver services, 
and concentrated populations of consuners 
of services, such as elder homes. 

Goals for 1994 include the development 
of modules in each of the areas of train- 
ing mentioned above. These modules will 
be available in multiple formats including 
audio tape, disk and large print. An 
outline of course content and audio 
description of overheads or slides will be 
a part of the module. 

FUTURE DIEECTIO® 

Other long term goals of the training 
program are to better utilize media such 
as satellite, cable T.V., radio networks, 
and other distance learning technologies. 

Training for professionals, both 
pre-service and inservice, is an ongoing 
process. Providing relevant and success- 
ful AT training in a cost effective manner 
remains a significant challenge for the 
field, and a major focus of many Tech 
Act programs. 



1 ) 



2 ) 



3) 



REFERENCES 

Rehabtech Associates, Inc. (1992). 
Training Programs in Assistive 
Technology. (Available from Rehabtech 
Associates, 3640 Dry Creek Court., 
Ellicott City, MD 21093) 

The Council for Exceptional Children- 
(1990). Resource Inventory; 
University-based Technology Training 
Programs. (Contract No. 300-87-0099) 
Washington, DC: U.S. Dept of ED. 

Peterson, D.B. , MacArthur, C.A. , 

Brady, M. (1992). Awareness & 
Information Needs About Assistive 
Devices in Maryland. [Summary] 
Proceedings of Resna International *92 
Conference 12, 179-181. 



ACKNOWL 






'JTS 



Funding for this project was provided by 
NIDRR Grant #224A90093. 



Nancy Friedman 
Training Director 
MD Technology Assistance Program 
MD Rehabilitation Center 
2301 Argonne Drive 
Baltimore, MD 21218 
(410) 554-3046 
Fax (410) 554-3206 



58 






RESNA ’94 • June 17-22, 1994 



44 




HOME SYSTEMS TECHNOLOGY FOR ELDERLY AND DISABLED PEOPLE 
PRESENT STATUS, R&D METHODOLOGY AND FUTURE DIRECTIONS 



Jos6 M. Ferreira Martyn Cooper and David Keating 

INESC - Largo Mompilher, 22 University of Reading • Whiteknights 

4200 Porto - PORTUGAL P. O. Box 225 - Berkshire R06 2AY - UK 



ABSTRACT 

The ability to perform everyday tasks, in the 
home, is a major need expressed by a large 
number of elderly and disabled people. The 
independent living that this leads to is aspired to 
by many. Although environmental control 
systems have traditionally been used to help them 
meet this need, several advantages can be 
identified if general home systems technology, 
with suitable user interfaces, is employed instead. 
This paper identifies the major issues in this 
approach, presents a brief overview of the state of 
this technology, discusses the requirements in 
terms of R&D methodology and outlines future 
directions in this area. 



BACKGROUND 

The needs of disabled and elderly people in the 
home are essentially the same as the needs of us 
all. However the way these needs are met may 
have to be different. A large number of disabled 
and elderly people need assistance for such 
common actions as switching on the lights, 
controlling white and brown goods, and many 
other simple tasks. This situation limits their 
independence in daily living, a problem that 
traditionally has been minimised by using 
environmental control systems. However, since 
each manufacturer develops its own proprietary 
solutions, adding new devices to the system is 
frequently not as simple as it should be. Lack of 
interoperability prevents higher manufacturing 
volumes and the limited scope of these systems 
make them not attractive to the general public 
(further contributing to increased prices). 

Market issues are indeed one of the most 
important questions that have to be addressed, 
since cost is usually the ultimate factor that 
determines whether or not a good product is 
successful in this under resourced sector. It is 
interesting to note that the single objective of the 
European TIDE (Technology Initiative for 
Disabled and Elderly people) programme has 
been stated as to "stimulate the creation of a 
single market in rehabilitation technology in 
Europe" [1]. Market issues in this area can only 
be successfully addressed if the manufacturers 
agree to work within defined standards, in order 



to assure equipment interoperability and low 
prices. Rehabilitation technology developed to 
improve independence at home can only meet 
these objectives if the wider technology domain 
of "home systems" is adapted to the specific 
requirements of disabled and elderly people. 

A Home System (HS) specification defines a 
"comprehensive communication system inter- 
connecting several kinds of electronic products 
used within the home" [2]. A commonly 
accepted, non-proprietary HS specification will 
ensure interoperability of equipment regardless of 
the manufacturer and at the same time will 
contribute to wider public acc^tance. TTie market 
for HS products has however consistently been 
under most expectations essentially due to a lack 
of an internationally accepted stantod [3]. This is 
still a largely unsolved issue, since different 
development directions were followed in J^an, 
US and Europe. 

Japan is probably the country where HS 
technology is more coherent, since the first 
activities were initiated as early as 1980 by the 
Institute of Electric Engineering of J^an [4]. A 
home bus study committee was set up in 1982 and 
the Home Bus System (HBS) standard was 
published in 1988. Over one million HBS 
standard systems are now installed in J^an. 

A less coherent approach is found in the US, 
where such different systems as the Smart House, 
CEBus and de-facto standards as X-10 and 
Echelon are available [3]. 

A number of specifications are also available in 
Europe, including Batibus and Instabus EIB [5]. 
The work towards a common European HS 
specification started in the late 80's, but the main 
effort started only in 1989 with an ESPRll' 
(European Strategic Programme for R&D on 
Information Technology) project entitled Home 
Systems (EP 2431), which released the first 
version Of the ESPRIT HS specification. 
Approximately one dozen ESPRIT projects are 
currently active in this domain, involving more 
than 50 companies developing Conformance and 
Development Tools as well as Software and 
Hardware Modules [6]. Version 1.1 of the HS 
specification was released in 1992 by the 
European Home Systems Association (EHSA), 
which was founded to promote and support the 
development and standardisation of this 
specification. 

59 



RESNA ’94 • June 17-22, 1994 



45 



HS Technology for Disabled and Elderly People 

STATEMENT OF THE PROBLEM 

General HS technology, according to the 
definition presented in the previous section, aims 
to achieve two main goals [7]: 

• To enable a service requiring several products 
to work together (which identifies the need for 
clusters of interconnected products) 

• To provide a simple way to operate 
sophisticated products (which is indeed one of 
the major consumer concerns) 

A HS able to fully meet these two requirements 
will provide three major benefits to the user: 

• It will allow remote control over much of the 
equipment within the house 

• It will allow several products to perform some 
type of co-ordinated operation or service 
(security, for example, may require the co- 
ordinated operation of different products, 
which may however be also used for other 
purposes) 

• It will make possible a true modular 
functionality, where products can be added or 
removed at any time 

Work done in this field in the last five years has 
produced the specifications of a flexible 
hierarchical architecture, and a communications 
network, which are able to provide these three 
benefits [2]. 

The practical advantages for the common 
consumer are of three main types: added comfort, 
improved security / safety and economy (mostly 
due to optimised energy consumption). However, 
for many disabled and elderly people, the 
important issue is that HS technology is enable 
of improving their independence in daily living, 
which definitely represents an important 
contribution to their qu^ty of life. Moreover, and 
since HS technology addresses the general 
consumer electronics market, it brings the 
additional benefits of low-cost and wide 
availability products. Allowing the community of 
disabled and elderly consumers to fully benefit 
from these promises can, however, be ensured 
only if two main requirements are satisfied: 

• The range of HS-compatible products 
available must be able to satisfy their specific 
requirements 

• The range of user interfaces available must be 
compatible with different types of disabilities 



APPROACH 



Moreover, and since these two requirements 
generate R&D tasks that can proceed in parallel, 
the approach to develop effective rehabilitation 
technology products in this area must include four 
main activities, in time order: 

• Identification of the end user requirements: 
The multi-disciplinary team responsible for 
this work must relate the requirements of their 
specific end us^ groups to all the information 
already available from other projects [8,9] 

• Technical specification: Again led by a multi- 
disciplinary team, this activity represents a 
first opportunity for system vs^dation, since 
rehabilitation technology professionals and 
end users are provided with a formal 
specification of the proposed solutions . 

• Applications and user interfaces development: 
This activity includes the two main R&D 
tasks corresponding to the requirements set up 
in the end of the last section. It is mainly an 
engineering led activity, although end user 
involvement is essential to provide guidance 
on many implementation details. Unnecessary 
duplication of R&D efforts can be assured by 
a proper survey of previous work, both in the 
gene^ HS technology field and in previous 
projects specifically concerned with the 
adaptation of HS technology to disabled and 
elderly users [10] 

• Evaluation: This activity comprises the 
implementation of solutions at end user sites, 
which must be accompanied by an adequate 
training program. A formal evaluation 
methodology will then enable the assessment 
of benefits [11] 

End user involvement is a key issue throughout 
all activities, since it is essential to guarantee that 
the developed solutions are effectively led by end 
user requirements and not by technology driven 
factors. In response to an open call for proposals 
set up during 1993 within the European TIDE 
programme, a two and a half year project 
following this ^)proach was selected for funding 
by the Commission of the European 
Communities. Started in January of 1994, this 
project is entitled HS-ADEPT (Home Systems - 
Access of Disabled and Elderly People to this 
Technology) and involves end user organisations, 
R&D institutions and industrial partners from 
four European countries. The workplan of the 
HS-ADEPT project closely foUows the four main 
activities previously described. 



The R&D methodology underlying any HS 
project aiming to meet the two requirements 
specified above can be summarised in one single 
statement: 

• Led by the end user’s requirements and not by 
the technology. 



O 



ERIC 46 



RESNA ’94 



IMPLICATIONS 



The implications of HS technology in the daily 
life of many groups of disabled and elderly 
people extend over a wide number of issues, the 
^ lyst important of which is however its very 






• Jane 17-22, 1994 



HS Technology for Disabled and Elderly People 

significant contribution towards greater 
independence. This is indeed a key issue, with 
further implications which range from the end 
users (improved self confidence, better quality of 
life, etc.) to the State itself (in terms of social 
support measures, financial issues, etc.). It is also 
important to note that the benefits of ad^ting HS 
technology to the specific requirements of 
disabled and elderly people, by making use of 
technologies and products addressing the wide 
consumer electronics market, effectively 
contributes to increase the awareness tow^ds the 
importance of a single market in rehabilitation 
technology products. 



DISCUSSION 

This paper has identified the niajor issues 
concerning HS technology for disabled and 
elderly people, presenting a brief overview of the 
state of this technology and discussing the 
requirements in terms of R&D methodology. It is 
however important to refer that HS technology, 
both for disabled and elderly people or for the 
general public, is still in its infancy. This is not 
only in the acceptance of internationally adopted 
standards, but also in the extraordinary potential 
that will result from the marriage with advanced 
telecommunication networks, namely the 
Broadband Integrated Services Digital Network 
(B-ISDN) [12]. This marriage will enable the 
development of a wide variety of telematic 
services, ranging from entertainment to education 
and tele-working. The multi-billion dollar market 
associated with sophisticated interactive video 
services will become a reality in the medium 
range future (10 to 20 yem say, depending on 
government regulations, world-wide 
standardisation efforts, etc.), but it is as yet 
unclear whether it will be possible to guarantee 
that this huge technological leap forward will 
contribute to yet further barriers for disabled and 
elderly people or be made accessible to them. 

It seems clear that it is possible to adapt HS 
technology to the specific requirements of people 
with special needs and still benefit from its 
success in the general consumer electronics 
market, but a large effort will certaiiily be 
required to make sure that the next step in tiiis 
(r)evolution will accommodate requirements 
other than those of the perceived mass markets. 



REFERENCES 

fl] Commission of the European Communities 
(DG XIII/C/3), TIDE 1993-1994 
Workplan, March 1993. 



[2] European Home Systems Association, 
Home Systems Specification, Release 1.1, 
March 1992. 

[3] P. Bord, "Home Automation International 
Environment," Actes de la Troisibme 
Confdrence Euro-Domotique, February 
1993, pp. 107-114. 

[4] K. Yamamoto, "Home Automation in 
Japan," Home Systems in a Global 
Environmental Workshop, November 1992. 

[5] T. Riley, "Emerging Standards in the 90's: 
A Strategy for Success," Actes de la 
Troisibme Conference Euro-Domotique, 
February 1993, pp. 81-82. 

[6] R. Torrenti, "The ESPRIT Home Systems 
Projects," Actes de la Troiiibme 
Conference Euro-Domotique, 1993, pp. TI- 
TS. 

[7] D. Fanshawe, "Resource Management in 
the Home System," Proceedings of the 
ESPRIT Conference, 1991, pp, 506-511. 

[8] D. Poulson and S. Richardson, "Issues in 
the Uptake of Adaptable Smarter Home 
Technology," Proceedings of the ECART 
Conference, May 1993. 

[9] R. Torrenti, "Home Systems and 
Handictqjped / Elderly People; A Necessary 
Coherent Approach," Proceedings of the 
First TIDE Congress, April 1993, pp. 157. 

[10] J. Falc6, J. Dolz, A. Mediano, J. Artigas 
and A. Roy, "Smart Homes for Disabled 
and Elderly: Adaptation and White Goods 
Modification," Proceedings of the ECART 
Conference, May 1993, 

[11] M. Lundman, "Methodology issues in R&D 
in rehabilitation technology," invited 
speech at the ECART Confaence, 1993. 

[12] A. SUenstra, M. Vaalen and J. Wage, 
"Telecommunications and the Introduction 
of Home Networks for the Residential 
Market," European Transactions on 
Telecommunications, Vol. 3, No. 1, 
February 1992, pp. 55-63. 



ACKNOWLEDGMENTS 

The authors are grateful to the Commission of *e 
European Communities for the funding which 
enabled the HS-ADEPT project. 

Jos6 M. Martins Ferreira 
INESC / University of Porto 
Largo Mompilher, 22 
4000 Porto - PORTUGAL 
Tel. 351-2-2094025 
Fax:351-2-318692 
E-maili jmmf@porto.inescn.pt 



61 



RESNA ’94 • June 17-22, 1994 



47 



ACCESSIBILITY STANDARDS FOR CmLDREN'S ENVIRONMENTS 



Bettye Rose Connell. Jon A. Sanford, Robin Moore, Jim Bostroni and Elaine OstrofT 

Center for Accessible Housing 
North Carolina State University 
Raleigh. NC 



ABSTRACT 



Despite obvious physical differences between 
ehildren and adults, child care centers, pre- 
kindergtuiens, elementoy schools, and other facilities 
used primarily by children are required to be built 
according to accessibility standards developed for an 
adult population. This paper describes the results of a 
projwt to identify the accessibility needs of children 
of different ages with a range of mobility aid visual 
impairments, and to evaluate the suitability of 
accessible features built to different technical 
specifications. Information was obtained through 
direct observation of children with disabilities using 
accessible features as well as through group 
interviews with children and adults. Accessibility 
needs were found to be dependent on personal 
factors, both developmental and disability-related, 
such as fine motor ability, upper body strength, extent 
of reach, stature, stamina, balance, and skill level in 
negotiating the environment, In addition, situational 
fwtors, such as the number of children with 
disabilities using a facility, policies and practices 
legaiding the role of adults in providing assistance, 
and intention of personal factors and physical 
characteristics of accessible features also were found 
to impact accessibility, 



BACKGROUND 



In child care centers, pre-kind^garten and elementary 
schools across rte country accessibility standards 
have been applied that were developed for an adult 
population. Although facilities complying with these 
standards have met the needs of many adult users 
with Usabilities working in or visiting children’s 
faciliries, such facilities have not necessarily met the 
access needs of children. In 1986, the Architectural 
and Transjwrtation Barriers Compliance Board 
(ATBCB) issued its own RecommnMons for 
Accessibility to Serve Physically Handicapped 
Children in Elementary Schools (1986). This was the 
first m^or step to address the lack of accessibility 
standards for children's environments, tailored to 
ehildren. In anticipation of the passage of the ADA 
and the need for more comprehensive guidelines for 
children’s environments, particularly those that were 
to be covered under Title II of the ADA (i,e., pubUc 
schools), the ATBCB contracted with the Center for 
Accessible Housing to develop recommendations for 
enhancing the accessibility of environments used by 
children of all ages, not just those of elementary . 



RESNA ’94 




school age and not just those with physical 
disabilities. The scope of these recommendations 
was to parallel and augment that addressed by 
ADAAGAJFAS for the adult population. 

OBJECTIVES 

The purpose of the project was to develop 
recommendations for accessibility guidelines for 
children's environments on the basis of: precedents 
established by existing codes and standards, findings 
of prior studies, and in-use evaluations of accessible 
children's environments. This paper reports on key 
factors impacting the suitability/accessibility of 
design features for children. 

METHOD 

Facility evaluations were conducted at nine sites that 
represented a variety of building types used regularly 
by children. These included: three preschools, three 
residential schools for children with vision and 
orthopedic disabilities, three elementary schools, and 
a children's museum. All had been built or renovated 
to be accessible. The sites, collectively, served 
children of diverse ages (e.g., preschoolers to teens) 
with different types of disabilities. 

The study focused primarily on children age 14 and 
younger. A total of 107 children were unable to walk 
and 44 walked with difficulty: 58 had no usable 
vision and 150 had low vision; 29 were unable to use 
their arms and/or hands, and 78 had limited use of 
their arms and/or hands; and 9 had severe hearing 
loss ^d 9 had mild to moderate hearing loss. In 
addition, many of the children had multiple physical 
and sensory impairments. The children used a wide 
variety of aids and devices to assist them in moving 
through the environment; however, substantial 
numbers of the children at several sites used no aid or 
device, at least under some circumstances. 

Physical measures data (e.g., slope and length of 
ramps, height of toilets) as well as illustrations and 
photographs were used to describe and document 
accessible features at each site. The suitability of as- 
built features for children with disabilities were 
assessed with a multi-method approach that involved 
observational techniques, self-reports of accessibility, 
and small group interviews. This approach 
evaluations permitted data to be obtained from 
multiple perspectives (e.g., self-report of children 
and/or staff regarding suitability of accessible 
features, observation by experts of suitability) that 
may or may not agree in tiieir assessments. 



wne 17-22, 1994 



48 



Childten's Environments 

Observational and self-report data were obtained on 
the suitability of frequently, infrequently and 
sensitive-use accessible features for children with 
disabilities. Observation of naturally-occurring use 
of accessible features was limited to those features 
that received a high incidence of use at predictable 
times. Simulations were utilized to obtain 
observational data for infrequently used and 
sensitive-use features, such as toilets. Children who 
participated in the simulated usage portion of the data 
collection w«e given simple instructions, such as 
’’show me how you would go about taWng a shower." 
Specific prompts to simulate use were not given 
unless the child had overlooked them. During the 
simulations, some children volunteered explanations 
of what they woe doing and why. In other cases, 
they answered questions posed by the resewch team 
related to the simulated activities. 

Small group interviews were the primary means of 
obtaining information from staff as to the suitability 
of accessible features for children as well as for 
themselves. Small group Interviews also were held 
with children when, in the judgment of the facility 
administrator, children could be identified that could 
meaningfully participate (e.g., old enough, verbal, 
etc.). The questions asked in the interviews were 
mo^fied, Spending on adult or child group 
composition. 

RESULTS 

Although the physical measures and site evaluations 
focused on a broad array of accessibility features that 
were likely to impact accessibility for children, those 
features that were found to be most critic^ to 
accessibility for children are highlighted in this paper. 
These include accessible routes, ramps, elevators, 
doors, storage and toilets. 

Arrfissihift Route: The width of accessible routes 
varied extensively within and across the sites (43" to 
138"). Where there were a number of children in 
wheelchairs who used a pathway at the same tinte, 
the wider routes worked well. They helped to avoid 
congestion and permitted two-way traffic. In 
addition, because children frequently do not have 
finely honed skills in maneuvering wheelchairs, the 
wider routes were compatible with their skill level, 
and reduced the wear and tear on walls and comers. 
Finally, in settings used by children in wheelchairs as 
well as children with sensory disorders, wider routes 
helped to prevent accidents. 

The length of interior and exterior accessible routes 
was also identified as an important issue because 
many children with disabilities at the study sites had 
limited stamina. Long accessible routes were 



common at multi-building sites, and in multi-story 
buildings. The distance that children had to fraverse 
also was Influenced by the placement of ramps and 
elevators in relation to the points of origin and 
destination for any specific olp. 

Ramns: Where ramps were present, they varied 
widely in slope. Five ramps were steeper than the 
widely accepted maximum slope of 1:12, ranging 
from 1:4 to 1:10. Four sites had ramps with slopes 
that were less than 1:12, ranging from 1:13 to 1:29. 
Many children, particularly those using manual 
wheelchairs who had limited strength and stamina, 
experienced difficulties ascending ramps, even those 
tlwt wwe shallow^ than 1:12. Ai a result, children 
were pushed up ramps by parents or staff at several 
sites. 

'The clear width of ramps also varied widely across 
the sites (36.5" to 116"). AU ramps exceeded the 
widely accepted minimum of 36". which helped to 
avoid congestion and accidents at sites where 
children in wheelchairs traveling in both directions at 
the same time. 

Eiftvatorsi Elevators were present at five sites, The 
^igTef eall buttons ranged from 40.5" to 43" above 
the finished floor (AFF). These heights were too 
high for many of the children to use easily, if at all. 
'Htese ^ficulties appeared to be relat^ to the body 
stature and reach of children. In addition, children 
with arm and hand impairments had even more 
difficulty in reaching controls. 

p ^rs: One of the most pervasive problems for 
children of different ages and with different types and 
severity of disabilities was pulling/pushing and 
holding doors open. Force required to operate 
extwior and interior doors varied widely. The least 
amount of force required to open an exterior door was 
2 pounds force (Ibf). More common force measures 
for exterior doors were between 8 Ibf and 16 Ibf. 
Some interior doors required as little as 1 Ibf to 3 Ibf 
U) open. Interior doors also were found that required 
as much as 16.5 Ibf to open. 

Virtually all children, not just those who use 
wheelchaira or those with gait and balance disorders, 
experienced difficulty in using doors. This was true 
for all doors requiring manual operation, although the 
amount of difficulty encountered appeared to be a 
function of the amount of force required and the 
strength of the child. The problem of force required 
to open exterior doors was avoided at two sites that 
used sliding doors with electronic/mechanical 
openers. At one preschool site a motion sensor 
activated a door opener, although the system could be 



iSiSST COi*¥ AVAILABLE 




63 

• Jnne 17^22, 1994 



RESNA *94 



49 



Children's Environments 

manually switched off by staff if necessary. At the 
other site, a power opener was used. These types of 
doors md openers easily accommodated age- and 
disability-related differences in the speed at which 
children could move through doorways. 

Sloraggi The maximum observed height for storage 
used by ambulatory children varied considerably, 
ranging ftom a low of 40" at a preschool to a high of 
78.5" AFF at a facility used by older children. 
Notably, storage 43" AFF was too high for older 
nonambulatory children at another site. The 40" 
height of the open storage cubbies at the preschool 
worked well for young children. At one site, some 
lockers had been modined with a low shelf 5" 
that worked well for children in wheelchairs with 
reach impairments. These data indicate that the most 
appropriate height of storage is influenced by body 
stature, body position, and forward and upward 
reaching capabilities of the child. It is unlikely that 
fixed storage heights will work for a cross section of 
children. Adjustable shelving of different depths is a 
viable option to this problem. 

loltets and Toilet Stalls; . Toilet seat height varied 
among the sites. The lowest toilets (11" to 12") were 
foimd at preschool sites. Sites that served older 
children and children of different ages and with 
diverse capabilities to transfer tended to have toilets 
of a variety of heights. Age and ind^endence in 
toileting influenced the suitability of different toilet 
seat heights. Staff reported that the critical issue 
related to toilet height for young children was to 
eiKure that they were low enough for children to sit 
with their feet on the floor, which enhanced Ae 
children's sense of stability and body functioning. As 
most young children in wheelchairs were placed on 
toilets, toilet heights compatible with transferring was 
less of an issue for this age group. At sites serving 
older children, toilets seats at 16" AFF permitted 
them to transfer independently. 

Grab bar height varied widely across sites (16" AFF 
to 42" AFF). Observations at several sites and staff 
comments suggest that the critical issue in the use of 
toilet side grab btus by children under 8 or 9 years old 
or those with disabili^-related reach limitations is the 
distance of the grab bar from the toilet, not its height. 
Different pab bar configurations worked well for 
children of different stature with different types and 
severity of orthopedic disabilities. In pre-school 
facilities, the use of small diameter bars enabled 
children to easily grasp them. 

DISCUSSION 

The suitability of accessible features at the study sites 
were impacted by differences among the children 



(and between children and adults) with regard to:(l) 
fine motor ability (such as the operation of controls), 
(2) upper body strength (such as opening doors), (3) 
reach in AP and ML planes (such as wall mount^ 
objects or grab bars), (4) stature (such as sitting on a 
toilet with feet on the floor), (5) stamina (such as 
climbing ramps), (6) balance (such as maintaining an 
upright position on a toilet), and (6) skill level in 
negotiating the environment (such as maneuvering a 
wheelchair precisely. In addition, the number of 
children with disabilities using a facility 
simultaneously, and the policies and practices of the 
institution with regard to encouraging independence 
impacted the suitability of accessible features in-use. 
These factors are important considerations in 
planning and implementing environmental and 
technologic^ changes that meet the accessibility 
needs of children with disabilities. An understanding 
of how these factors shape accessibility needs and 
experiences is crucial in the development of 
technological and environmental solutions to 
accessibility in children's environments. 



REFERENCES 

Architectural and Transportation Barriers 
Compliance Board. (1986). Recommendations for 
Accessibility to Serve Physically Handicapped 
Children in Elementary Schools, Washington, DC. 



ACKNOWLEDGMENTS 

This project was completed under a contract from the 
U.S. Architectmal and Transportation Barriers 
Compliance Board. 



Bettye Rose Connell, Ph.D. 

Director, Resetuch and Design Development 
CentCT for Accessible Housing 
Box 8613 
NCSU 

Raleigh, NC 27695-8613 

Phone 404/728-4840 FAX 404/728-773 1 




50 



RESNA ’94 • June 17-22, 1994 



SIG-02 

Personal Transportation 



Development of instrumentation and protocols to measure 
the dynamic environment of a modified van 

Maureen Linden, MS Transportation REC, University of Virginia 
Stephen Sprigle, PhD, Center for Assistive Technology, SUNY-Buflfalo 



Abstract. T he dynamic environment of a van modi* 
fied to accommodate a person driving Srom a 
wheelchair was measured to determine the effects 
of position within the van and the type of seat used. 
The project measured accelerations as a subject sat 
in 3 positions within the van and upon two different 
seats. Three separate van maneuvers at different 
speeds were used to change the dynamic environ- 
ment. Van accelerations at the different positions 
varied significantly. A wheelchair transmitted more 
accelerations to the subject than the original equip- 
ment manufacturers (OEM) seat, making it harder 
to maintain a stable posture. These results should 
prove useful to others studying the functional abili- 
ties of wheelchair users within a vehicle 
environment. 

Introduction . In order to determine the 
effects of the dynamic environment on the stability 
and functional ability of a driver with a disability, 
the forces and accelerations within the driving envi- 
ronment must be understood. A person driving from 
a wheelchair usually requires modifications to a 
van, including a raised roof and dropped floor, and 
a lift. These modifications influence the vehicle's 
handling characteristics. 

Much of the research that has been per- 
formed on dynamic vehicular environments focused 
on vehicles during high performance or emergency 
maneuvers. Very limited information is available 
regarding the dynamics of vehicles after they have 
been modified to accommodate a person seated in a 
wheelchair. Literature on human reactions in 
moving vehicles deals almost entirely with able- 
bodied individuals or crash test dummies seated in 
the original manufacturer's seating systems. 

Objective. The goal of this project was to 
design instrumentation and testing protocols which 
could be used to defme the dynamic environment 
presented by a vehicle which had been modified to 
accommodate wheelchair users. This study concen- 
trated on identifying the effects of seat type and seat 
position within the vehicle, and quantifying the 
magnitudes and variation of accelerations during 
repeated trials. Information collected during this 
project could then be used to design studies which ^ 



specifically study the effect of the wheelchair, seat- 
ing system, and/or van modifications on the stability 
and function of a wheelchair user in a vehicle. 

Methods. The response to the dynamic 
envirorunent of an able-bodied subject was charac- 
terized by the linear accelerations as a function of 
her position within the van and the seat upon which 
she was seated. Accelerations were measured with 
the subject seated in the manufacturer's driver seat, 
the front passenger seat, and a wheelchair secured 
in the driver's and rear passenger areas. Subject 
response was measured by accelerometers secured 
to fre sternum of the subject. Vehicle response was 
measured by accelerometers secured under the seat 
of the subject. Acceleration profiles of the vehicle 
and subject were collected while the vehicle 
performed several maneuvers outlined below. 
Maneuvers were selected because they represented 
common tasks in a normal driving experience. 

1. Acceleration from rest Van was accelerated 
from standstill to speeds of 16.1, 32.2, and 48.3 
km/hr (10, 20, and 30 mph) in ten seconds. 

2. Deceleration to rest. Van was decelerated 
from speeds of 16.1, 32.2, and 48.3 km/h (10, 20, 
and 30 miles/hour) to standstill in 4 secs. 

3. Curve Driving. Van was driven around a curve 
at constant speeds of 16. 1 and 32.2 km/hr. 

Data Collection and Analysis . Accelera- 
tions were measured using 6 uniaxial accelerome- 
ters which had a range of +/- 5 g and a frequency 
range of .25 to 10^ Hz. Data was sampled by a 16 
bit A/D converter and stored on magnetic tape. The 
output of each accelerometer was filtered using a 
10th order low pass filter with a cut off frequency 
of 1 Hz. The resultant acceleration vector was cal- 
culated as a function of time given the 3 orthogonal 
accelerations of the subject and vehicle. A root- 
mean-square (RMS) value was calculated to 
represent the amount of acceleration present in each 
maneuver. For the acceleration and deceleration 
maneuvers, RMS of the total and longitudinal 
accelerations were calculated, and for curve data, 
the lateral and total RMS was determined. The 
respective imiaxial accelerations were considered 



66 

RESNA ’94 • June 17-22, 1994 



52 



Measurement of dynamic van environment 

the primaiy accelerations for the maneuvers. 

A two-factor ANOVA was performed to 
determine the effects of position within the vehicle 
and the speed of the maneuver. Subsequent analysis 
also considered the effect of sitting in an OEM seat 
or a wheelchair. 

Results/Discussion. 

Position in Vehicle: Comparison of the (fynamics at 
different locations within the vehicle showed differ- 
ences between the rear passenger area and the front 
of the vehicle. The front passenger and driver areas 
experienced comparable average RMS accelera- 
tions, with the front passenger area averaging 1 .4 
rimfig the driver area (Table 1). The rear passenger 
RMS accelerations were the greatest, averaging 
almost 3 times those of the driver area (Table 1). 
The effect of position on vehicle dynamics was sta- 
tistically significant for primary and total accelera- 
tions during accelerations and deceleration maneu- 
vers, but not during curve maneuvers (Table 2). 
During deceleration maneuvers, the effect of the 
position was dependent on the speed from which the 
vehicle decelerated. The van pitched during the 20 
and 30 mph decelerations, but was not as noticeable 
during the 10 mph decelerations. As this pitching 
became more severe at higher speeds, its effects on 
the linear accelerations present at different 
locations within the vehicle changed. 

The subject exhibited similar accelerations 



in the front passenger and driver location, except 
during acceleration maneuvers (Table 1). For these 
maneuvers the driver showed a total acceleration 
an order of magnitude higher than the front passen- 
gGT, even though the longitudinal accelerations were 
similar. This could be partly because the subject in 
the front passenger position was sitting passively, 
while the driver subject was actively controlling the 
vehicle. In addition to grasping the steering wheel, 
the active participation in driving could alter the 
torso accelerations when compared with a subject 
who is sitting passively. 

The subject seated in a wheelchair in the 
rear passenger area of the vehicle experienced simi- 
lar RMS accelerations in a wheelchair secured in 
the driver position (Table 2). The total and lateral 
RMS accelerations were greater for the rear pas- 
senger area during deceleration maneuvers and 
similar during acceleration maneuvers. During the 
curve maneuvers, the lateral and total accelerations 
were less in the rear location than in the driver area. 

The results of these comparisons might 
influence research into driver evaluation. The dif- 
ferences between the dynamics at different positions 
within the vehicle should be taken into careful con- 
sideration when evaluating or studying the stability 
and function of a driving candidate. While the ideal 
situation would be to allow the individual to experi- 
ence the dynamics of the vehicle from the driver's 



Table 1: Average g RMS Comparison of Vehicle Positions 



Maneuver 


Vehicle 


Subject 


Vehicle 


Sul 


bject 






Driver 


Rear 


Driver 


Rear 


Driver 


Front 


Driver 


Front 






in w/c 


Pass 


in w/c 


Pass 


in 


Pass in 


in 


Pass in 








in w/c 




in w/c 


OEM 


OEM 


OEM 


OEM 


accel 10 


longitud 


0.0338 


0.0806 


0.0291 


0.027 


0.0338 


0.0352 


0.0178 


0.0227 




total 


0.051 


0.155 


0.0481 


0.0427 


0.051 


0.0375 


0.165 


0.0226 


accel 20 


longitud 


0.0447 


0.129 


0.0631 


0.0224 


0.0447 


0.0508 


0.021 


0.0316 




total 


0.157 


0.18 


0.0879 


0.0425 


0.157 


0.0699 


0.187 


0.0807 


accel 30 


longitud 


0.0402 


0.114 


0.0961 


0.134 


0.0402 


0.121 


0.0358 


0.0299 




total 


0.123 


0.176 


0.143 


0.248 


0.123 


0.26 


0.183 


0.0438 


decel 10 


longitud 


0.0247 


0.117 


0.0716 


0.0961 


0.0247 


0.0937 


0.0356 


0.0488 




total 


0.0579 


0.167 


0.144 


0.119 


0.0579 


0.11 


0.0573 


0.0677 


deoel 20 


longitud 


0.071 


0.173 


0.203 


0.25 


0.071 


0.107 


0.0941 


0.113 




total 


0.131 


0.308 


0.265 


0.364 


0.131 


0.161 


0.101 


0.141 


decel 30 


longitud 


0.12 


0.397 


0.307 


0.358 


0.12 


0.0973 


0.0798 


0.156 




total 


0.211 


0.657 


0.428 


0.51 


0.211 


0.183 


0.0876 


0.194 


curve 10 


lateral 


0.0557 


0.0913 


0.178 


0.082 


0.0557 


0.0559 


0.0343 


0.0545 




total 


0.0828 


0.359 


0.326 


0.162 


0.0828 


0.0127 


0.0517 


0.0851 


curve 20 


lateral 


0.0652 


0.201 


0.128 


0.0308 


0.0652 


0.0638 


0.0527 


0.122 




total 


0.254 


0.382 


0.378 


0.251 


0.254 


0.114 


0.0742 


0.145 



RESNA ’94 



67 

June 17-22, 1994 



53 



Measurement of dynamic van environment 



location, this might not be appropriate. Similar 
dynamics of the driver and fiont passenger areas 
demonstrate that the abilities of the driving 
candidate in the front passenger area would be in- 
dicative of his/her ability in the driver area. The 
significantly higher accelerations in the rear pas- 
senger area show that performance does not 
necessarily represent performance in the driving 
task, but because it is a more severe environment, 
evaluation of a candidate in diis position could be 
useful in driver training and evaluation. 

OEM seat vs. wheelchair. The wheelchair 
transmitted 2.13 times more longitudinal accelera- 
tion than the OEM seat during the acceleration and 
deceleration maneuvers (Table 4). During the de- 
celeration maneuvers, the wheelchair transmitted 
3.7 times more total RMS acceleration, while for 
accelerations the OEM seat transmitted 2.7 times 
more total RMS acceleration. During the curve ma- 
neuvers, the wheelchair transmitted 4.7 times more 
total and lateral RMS acceleration. The greater ac- 
celerations during deceleration and curve maneu- 
vers makes it a less suitable driving seat than the 
OEM seat because these maneuvers are typically 
more unstable than acceleration maneuvers during 



which the seat back provides support. 

Conclusion . This project began to defme 
the complex dynamic environment within a 
modified van. This environment affects the 
functional ability of a person with a disability, and 
a clearer understanding of it will permit better 
research and clinical practice in evaluating the driv- 
ing ability of persons with disabilities. 

The accelerations experienced by a person 
in a moving van was found to be dependent on both 
the position within the vehicle and the seat upon 
which a person is seated. The OEM seat dampens 
accelera^ons and provides a more stable seat. All 
of the data in this study was taken in the same vehi- 
cle. Further research should use vehicles with dif- 
ferent modifications to determine the effects of 
individual modifications on the vehicle (tynamics. 

Acknowledement . This research was 
performed at the Transportation Rehabilitation 
Engineering Center at &e University of Virginia, 
and was funded by NIDRR Grant# H133E00006 
For further information, contact: 
Transportation REC, 1011 Linden Ave, 
Charlottesville, VA 22901, 804/296-7288 



Table 2 

Maneuver 

Acceleration 

Deceleration 



Primary axis acceleration 



Total acceleration 



speed effect significant 
position effect significant 
speed X position interaction 
significant 



position effect significant 
speed X position interaction 
significant 



Table 3: Average g RMS Comparison of Wheelchair vs. OEM Seat tDiiver Position) 


Maneuver 




Subject 


OEM Seat 


Wheelchair 


accel 10 


longitudinal 


0.0178 


0.0291 


total 


0.165 


0.0481 


accel 20 


longitudinal 


0.021 


0.0631 


total 


0.187 


0.0879 


accel 30 


longitudinal 


0.0358 


0.0% 1 


total 


0.183 


0.143 


decel 10 


longitudinal 


0.0356 


0.00716 


total 


0.0573 


0.114 


1 decel 20 


longitudinal 


0.0941 


0.203 


total 


0.101 


0.265 


decel 30 


longitudinal 


0.0798 


0.307 


total 


0.0876 


0.428 


curve 10* 


lateral 


0.0343 


0.178 


total 


0.0517 


0.326 


1 curve 20 


lateral 


0.0527 


0.128 


total 


0.0742 


0.378 



u :;B8 

RESNA ’94 • June 17-22, 1994 



54 



Accelerations experienced by wheelchair users with SCI in a moving van 



Stephen Sprigle, PhD, Center for Assistive Technology, SUNY-BufFalo 
Maureen Linden, MS, Transportation REC, University of Virginia 



O 

ERIC 



Abstract. W heelchair users who drive must over- 
come the dynamic vehicle environment to maintain 
stability. Proper stability might increase function 
and improve the driving ability of wheelchair users. 
This project measured the accelerations experienced 
by individuals with SCI during vehicle maneuvers. 
Ihe subjects sat both passively and while grasping 
a structure to simulate gripping driving controls. 
Seated stability was greater when the subjects 
grasped the structure. Vehicle and subject responses 
vari^ greatly. This result underscores the complex- 
ity involved when researching the functional 
abilities of a SCI population within a complex 
dynamic environment. 

Introduction. Some wheelchair users are unable to 
transfer into a car and, therefore, must enter a van 
while seated in a wheelchair. Once in the van, 
some transfer to a typical van seat, while others 
drive their vehicles while seated in their wheel- 
chairs. Typically, vans are modified in two ways to 
accommodate the increased sitting height of a per- 
son in a wheelchair, the floor of the van is lowered, 
a raised roof is added, and a wheelchair lift is 
mounted to the rear side door. These modifications 
change the vehicle (tynamics by changing such 
properties as mass, mass distribution, and center of 
gravity. 

Seated stability is compromised by the 
dynamic driving environment which affects an indi- 
vidual's ability to maintain a stable posture 
(Kulowski, 1960). While seated in a moving vehi- 
cle, the musculature of the trunk, legs, and hips are 
constantly adjusting to the dynamic environment to 
maintain an upright posture. For the able-bodied 
population, this is an unconscious response. For a 
person with a spinal cord injury, the maintenance 
of driving posture can become a conscious task. 

Posture while driving is stabilized by 
providing the driver with a large base of support, 
through use of the seat, backrest, and terminal 
extremity contacts. Grasping the driving controls 
closes the links created by the hands gripping the 
steering control and the shoulder contact with the 
backrest (Zacharkow, 1988). 

Objective . This study positioned persons 



with cervical level spinal injuries in a wheelchair to 
measure the (fynamic response of the trunk within a 
moving vehicle. Measurements were taken while 
the subjects sat passively and while gripping a 
structure. Accelerations at the sternum and a meas- 
ure of trunk position was used to relate vehicle 
maneuvers to trunk accelerations and stability in 
persons with SCI. The study was designed to be a 
launching point for an extensive quantification of 
the driving environment and its effect on the func- 
tional ability of a person with a disability. 

Instrumentation . The relative position be- 
tween the subject's trunk and the wheelchair back- 
rest was measured using plate switches. Five 
switches oriented in a T configuration, a row 
mounted along the top edge of the upholstery, and 
two additional switches placed vertically along the 
spine. A rough measure oftrunk position was 
determined by recording whether a switch was open 
or closed. 

Accelerations were measured with six um- 

= +/-5g; 
accelerome- 
nting cubes. 

Vehicle response was measured by accelerometers 
mn untPil on a cube which was secured to the floor 
of the vehicle and centered between the casters. 
Subject response was measured by accelerometers 
mounted on an identical cube secured two to three 
centimeters inferior to the stemo-clavicular joint of 
the subject. 

Methods . Subjects were four adult drivers 
and non-drivers with spinal injury levels fiom C4 to 
C6. The responses of the subjects were measured 
with each subject seated in a standard adult manual 
wheelchair secured to the van using an four-point 
belt tie-down. 

Vehicle maneuvers were performed while 
the subjects were seated with hands in their laps, 
and again, while they grasped a structure mounted 
to the floor in front of the wheelchair. The structure 
was mounted to the floor in fiont of the subject and 
adjusted to allow each subject to grasp it as they 
would grasp a spirmer knob and hand controls. 

Acceleration profiles of the vehicle and 

■69 



axial piezobeam accelerome^ (range 
firequency range: 0.25 to 10 Hz). The 
ters were mounted on two tri-axral mou 



RESNA ’94 • June 17-22, 1994 



55 



SCI response in van 



subject, and general positioning of the subject 
relative to the seat back were collected while the 
vehicle performed several maneuvers outlined 
below: 

1. Acceleration from rest. Van was accelerated 
from standstill to speeds of 16.1, 32.2, and 48.3 
km/hr (10, 20, and 30 mph) in ten seconds. 

2. Deceleration to rest. Van was decelerated 
from speeds of 16.1, 32.2, and 48.3 km/h (10, 20, 
and 30 miles/hour) to standstill in 4 seconds. 

3. Curve Driving. Van was driven around a curve 
at constant speeds of 16. 1 and 32.2 kmyhr. 

Data Collection and Analysis . The output 
of each accelerometer was filtered using a 10th 
order low pass filter with a cut off fi^quency of 1 
Hz. The resultant acceleration vector was cal- 
culated as a fimction of time given the 3 orthogonal 
accelerations of the subject and vehicle. A root- 
mean-square (RMS) value was calculated to 
represent the amount of acceleration present in each 
maneuver. 

The effects of the stabilizing structure on 
the stability of the subjects was expressed by the 
relationship between the acceleration of the subject 
and the vehicle. A transmission ratio was defined as 
the ratio of the RMS acceleration of the subject to 
the RMS acceleration of the vehicle. 



Another estimate of the stabilizing effect of 
the structure was made using the total time switches 
were open (Tqj) during maneuvers. The duration of 
each maneuver was calculated. T^g was used to de- 
termine the normalized percentage of time switches 
were open during each maneuver for each subject. 

Results. Without the structure, all subjects 
showed and increase in Tos during deceleration and 
curve maneuvers (Table 1). An increase in the total 
RMS subject acceleration yielded a decrease in Tq^ 
during acceleration maneuvers and an increase dur- 
ing decelerations and curves. 

No significant relationship was found 
between the total RMS vehicle acceleration and the 
subject transmission ratio because of the wide 
variation in vehicle RMS (Table 2). 

Discussion . A total of four subjects with 
spinal cord injuiy were used in this study , and 
despite the fact that all are wheelchair users who 
use or would use a van, the differences in stability 
and trunk movement were extreme. Two subjects 
reported using the structure to stabilize their trunks, 
and the effect of the structure in limiting trunk 
movement was evident. Research into driver 
performance should, therefore, use a structure to 
simulate the closed-link stability normally provided 
in driving. The fimctional ability of a person in a 



Tabid; Normalized Switch Open Times 









No Structure 


Structure 


Subject 


Maneuver 


Max Possible 
T„s (secs) 


% Total 

1'ns 


% Total 
Tn« 


LM 


accel 10 


44 


41.6 


25.7 




accel 20 




37 


25.7 




accel 30 




28 


25.7 




decel 10 


20 


35.3 


12.7 




decel 20 




73.9 


24 




decel 30 




89.7 


23.8 




curve 10 


60 


47.8 


25 




curve 20 


40 


71.8 


19.1 


JF 


accel 10 


44 


0 


0 




accel 20 




0 


0 




accel 30 




0 


0 




decel 10 


20 


r 0.3 


0 




decel 20 




11.4 


0 




decel 30 




38.8 


0 




curve 10 


60 


0 


0 




curve 20 


40 


16.8 


0 


MT 


accel 10 


44 


0.5 


0 




accel 20 




2.8 


2.3 




curve 20 


40 


14.8 


20.8 



7 0 RESNA ’94 • June 17-22, 1994 



56 



SCI response in van 

moving vehicle would be uffected if subjects were 
allowed to stabilize their trunk with their upper 
extremities. The concern of this study was to define 
what happens in this environment. Despite the fact 
that using one's upper extremities for stability might 
not be proper, it should be taken into account wiien 
studying this population. 

Studjong the accelerations at the sternum 
of subjects with SCI and the subsequent movement 
of their trunks, provided a preliminary look at the 
variations of response to the dynamic environment. 
Variations in vehicle and subject RMS accelera- 
tions, and the variable effect on subject RMS 
caus^ by the structure more clearly defines the 
complexities involved in studying this environment 
using persons with SCI. 

The results can be applied by researchers 
to simulate the driving environment with a 



simulator or other vehicle. Using a dynaimc envi- 
ronment in a preliminary driving evaluation would 
allow evaluators to determine if a driving candidate 
is stable enough to withstand the driving environ- 
ment, and determine which adaptive driving con- 
trols are the most appropriate, without placing the 
individual behind the wheel of a moving vehicle. 
References 

• Kulowski J, Crash Injuries, Springfield IL: 

Charles C Thomas, 1960 

• 2[acharkow D, Sitting, Standing, Chair Design, & 
Exercise. Springfield, IL Charles C. Thomas, 1988. 

A cknowledement . This work was 
performed at the Transportation REC at the 
University of Virginia and was funded by NIDRR 
grant #H133E00006. 

Further information : contact 
Transportation REC, 1011 Linden Ave. 
CharlottesviUe,VA 22901,804-296-7288 



Maneuver 


Subject 


Structure No Structure 


Vrms 


SrmsA^rms 


Vrms 


SrmsA^rms 


Accel 10 


RP 


.0763 


0.293 








JF 


.24 


0.601 


.0837 


0.351 




LM 


.0689 


0.608 


.216 


0.26 




MT 


.161 


0.81 


.126 


0.911 


Accel 20 


RP 


.147 


0.246 








LM 


.107 


0.691 


.116 


0.559 




JF 


.406 


0.196 


.114 


0.621 




MT 


.171 


0.521 


.235 


0.595 


Accel 30 


RP 


.125 


0.205 








LM 


.125 


0.482 


.155 


0.75 




JF 






.251 


0.452 




MT 






.19 


0.926 


Decel 10 


RP 


.072 


0.372 








LM 


.0902 


2.663 


.158 


0.799 




JF 


.385 


0.363 


.0978 


1.007 


Decel 20 


RP 


.172 


0.72 


.063 


0.381 




LM 


.183 


0.315 


.315 


0.608 




JF 


.728 








Decel 30 


RP 


.281 


0.72 








LM 


.33 


0.457 


.144 


6.504 




JF 


1.47 


0.199 


.72 


0.454 


Curve 10 


RP 


.131 


0.305 








LM 


.229 


0.338 


.133 


0.535 




JF 


.673 


0.188 


.673 


0.188 


Curve 20 


RP 


.403 


0.0958 








LM 


.161 


0.586 


.268 


0.566 




JF 


.363 


0.646 


.71 


0.264 




MT 


.601 


0.219 


.389 


0.519 



er|c 



RESNA’94 • June 17-22, 1994 



57 



ROLLOVER AND DIRECTIONAL STABILITY OF VEHICLES MODIFIED FOR 

THE PHYSICALLY CHALLENGED 

W. D. Pilkey, J. R. Brazell, W. Kang, and L. Kitis 
Department of Mechanical Aerospace, and Nuclear Engineering 
University of Virginia 
Charlottesville, VA, U.S.A. 



Abstract 

A new dynamic rollover stability criterion for 
tripped rollover, the mini mum lateral impulse, is 
defined. This criterion and computer simulations of 
vehicle motion during a rollover accident are used to 
compare the rollover stability of an original 
manufactured van with three vans modified to 
accommodate physically challenged passengers or 
drivers. 

Introduction 

Standard manufactured vehicles are often subjected to 
certain structural changes designed to accommodate 
physically challenged passengers or drivers. The main 
objective of the research reported in this paper is to 
study the effect of such adaptive changes on the 
rollover stability of the vehicle. Specifically, 
unmodified, or OEM (Original Equipment 
Manufacturer) vans, are conq)ared with vans 
modified as described below: 

a) Mod B - the roof of the van is made 13 inches 
higher and a wheelchair lift is added. 

b) Mod C - the floor of the van is lowered 6 inches, 
its body is lifted 2 inches, and a wheelchair lift is 
added. 

c) Mod D - the floor is lowered 4 inches and a 
wheelchair lift is added. 

There are two types of rollover situations that are of 
interest; tripped and untripped. A tripped rollover 
occurs when the vehicle is skidding sideways, and 
strikes a tripping object such as an obstacle, a curb, 
soft soil, or other similar terrain feature. The 
untripped or maneuver- induced rollover results from 
abrupt maneuvers or excessive driver inputs, such as 
oversteering, or high speed turning or cornering. 

A large majority of single vehicle rollover accidents 
are classified as tripped rollovers. According to 
NHTSA’s reports, 90 percent of all rollover accidents 
are single car accidents which occur off the roadway. 
This inq)lies that most rollover accidents, which are 
caused by a loss of directional control, are tripped 
rollovers. Thus, although the parameters involved in 
a tripped rollover analysis are independent of 
maneuvering, the roll propensity of a given vehicle is 

:.72 



related to its directional or handling stability. 

The National Highway Traffic Safety Administration 
(NHTSA) has proposed three measures of vehicle 
stability that quantify the roll propensity of a vehicle: 
fhg st atic stability factor (SSF), the tilt table ratio 
(TTR), and the side pull ratio (SPR). The static 
stability factor SSF is defined as the ratio of half 
track width to the vertical height of the center of 
gravity. The tilt table ratio TTR is measured with the 
vehicle placed on a platform which is tilted about an 
axis parallel to the longitudinal axis of the vehicle. 
Rollover stability is assumed to be characterized by 
the angle at which the tires on the upper side of the 
platform begin to lose contact with the platform. The 
numerical value of the TTR is defined as the tangent 
of this angle. The side pull ratio SPR is measured by 
pulling the vehicle in the lateral direction at the 
height of the center of gravity until the tires on the 
far side of the vehicle lose contact with the ground. 
The SPR is defined as the ratio of the lateral force to 
the weight of the vehicle. These simple static criteria 
have been applied to both tripped and untripped 
rollovers. 

Dynamic Criteria for Rollover Stohility 

The SSF, TTR, and SPR are all static criteria. 
Because the rollover process involves a complex 
interaction of forces from suspension systems, tires, 
drivetrains and road surface, dynamic criteria are 
needed for predicting rollover propensity under 
dynamic conditions. It is possible to conclude that a 
vehicle is more stable than another vehicle according 
to a static metric and find that under certain dynamic 
conditions the vehicle becomes less stable. This can 
happen when the frequency content of the lateral 
forces or accelerations includes some of the natural 
frequencies of the vehicle. The two existing dynamic 
criteria for rollover stability are the critical rollover 
speed and the critical rollover acceleration, which are 
defined as the least lateral speed and the least lateral 
acceleration required for rollover. These two criteria 
are dependent on parameters external to the vehicle, 
such as the properties of the tripping device in 
tripped rollover , and, consequently , are not 
determined only by the structural properties of the 
vehicle. 



RESNA ’94 • June 17-22, 1994 



58 



ROLLOVER AND DIRECTIONAL STABILITY 



This paper introduces a new dynamic criterion called 
minimum lateral impulse (MLI) for measuring vehicle 
tripped rollover stability[l]. The MLI criterion uses 
as a measure of rollover stability the minimum 
impulse in the lateral direction over the period of 
time from the start of the rollover process to the 
instant when the vehicle is just on the verge of 
overturning. The minimization of this impulse is with 
respect to all possible time histories of lateral 
deceleration over the time interval of interest. Thus, 
the lateral deceleration function for which the lateral 
impulse assumes its least magnitude while still 
causing rollover is to be determined. Physically 
meaningful constraints, such as upper and lower 
bound bands on the allowable lateral deceleration 
time histories, can be included in the minimization. 
The differential equations governing the motion of the 
vehicle are also treated as constraints. Thus, rollover 
stability analysis with the MLI criterion is formulated 
as a constrained optimization problem, which can be 
solved by mathematical programming methods. The 
numerical value of the MLI measure is usually given 
per unit mass of the vehicle so that an equivalent 
definition of this criterion is the time integral of the 
lateral acceleration from the onset of rollover to the 
moment just before the vehicle overturns. Hence, the 
value of the MU measure has dimensions of velocity, 
m/s or ft/sec. 

Parametric computer simulation studies can also 
account for dynamic effects in rollover stability. The 
Vehicle Dynamic Analysis Nonlinear (VDANL) 
simulation program[2] is currently being used to 
study the directional and roll stability of modified 
vans. 

Results 

Figure 1 shows a simplified six degree-of-freedom 
vehicle model, m, and m„ are the masses of the 
sprung and unsprung parts of the vehicle and the 
coordinates y^, Zj, and $| , i = u, s, are the degrees 
of freedom. This model was used to represent the 
OEM van and the three modified vans, Mod B, C, 
and D, in an MLI analysis. Figure 2 shows the 
deceleration curves that minimize the lateral impulse 
and the corresponding numerical values of the MLI 
measure for each of these four vans to reach their 
rollover angle within 1 second in tripped rollover. 
The relevant design variables and a comparison of the 
MU method to the SSF method is shown in Table 1. 
The static stability criterion SSF indicates that the 
most stable van is the OEM van and the least stable 
van is the Mod B van. According to the dynamic 
stability criterion MLI, however, the most stable van. 



for which the lateral impulse required to tip the 
vehicle over is the greatest, is the Mod B van and the 
least stable van is the OEM van. The reason for this 
apparent discrepancy is that the SSF metric depends 
only on the height of the center of gravity and the 
track width of the vehicle, while the MLI criterion 
accounts also for the increase in the sprung mass M. 
and, more importantly, for the change in the mass 
moment of inertia I„. Table 1 shows that all three 
modified vans have a greater I^ than the OEM van. 
Therefore, a greater moment and a greater MU is 
required to bring the vehicle to its rollover angle. For 
the vehicle modifications studied here, the increase in 
I„ more than offsets the effect of increases in center 
of gravity height and sprung mass. 

Conclusion 

The rollover stabilities of an OEM van and three vans 
modified for use by the physically challenged have 
been studied. It has been found that, in tripped 
rollover, the new dynamic criterion MLI for rollover 
stability predicts different results from those obtained 
by static criteria. 

References 

1. Pilkey, W. D., Kang, W., and Kitis, L., "A 
Stability Criterion for Vdiicles in Tripped Rollover", 
Proceedings of the 64th Shock Vibration 
Symposium, 25-28 October 1993, Florida. 

2. Allen, R. W., etc. Vehicle Dynamic Stability and 

Rollover, Report No. STI-TR-1 268-1 , System 

Technology, Inc., June, 1992. 

Acknowledgments 

This research was sponsored by the National Institute 
on Disability and Rehabilitation Research (NIDRR) 
and the Virginia Center for Innovative Technology. 




Fig. 1 Vdiicle model 




RESNA’94 • Jane 17-22, 1994 



59 



ROLLOVER AND DIRECTIONAL STABILITY 



Table 1 Inertial Parameters and Rollover Stability Criteria of Four Ford B150 Vans 



VAN 


M. 

0b-McVft) 


c.o. 

(ft) 


L 

(Ib-ft-wc*) 


SSP 


MLI 

(ft/scc*) 


OBM 


136.0 


2.54 


578.8 


1.115 


11.99 


ModB 


146.4 


2.72 


2144.0 


1.042 


15.23 


Mode 


142.0 


2.67 


1028.0 


1.063 


12.82 


ModD 


142.0 


2.50 


878.0 


1.100 


12.99 




Time-sec 




Time-sec 





Time— sec 



Fig. 2 Minimized deceleration pulses for the four vans 



Walter D. Pilkey 
Morse Professor 

Department of Mechanical Aerospace and Nuclear 

Engineering 

University of Virginia 

Charlottesville, VA 22903 



BEST COPY AVAILABLE 74 




RESNA ’94 • June 17-22, 1994 



CRASH RESPONSE OF WHEELCHAIR-OCCUPANT SYSTEMS IN TRANSPORT 



W. D. Pilkey, W. Kang, and G. Shaw 
Transportation RBC, University of Virginia 
Charlottesville, VA, U.S.A. 



Abstract 

A computer model has been developed to simulate the 
crash effects on a wheelchair-occupant system 
secured to the floor of a vehicle. The DYNAMAN 
simulation package has been used to simulate the 
system that includes a wheelchair surrogate and a 
Hybrid III dummy. The simulation results agree well 
with the results of three laboratory sled tests. It has 
been verified that this model can be used to predict 
and corroborate sled test results in inter-laboratory 
tests. 

Introduction 

The increasing need of mobility for the physically 
challenged results in a significant driver and 
passenger population who sit in wheelchairs in vans 
and buses. The wheelchair tiedowns and occupant 
restraint systems can reduce the possibility of injury 
by preventing the wheelchair and occupant from 
moving and hitting the vehicle interior during a 
crash. To develop a wheelchair tiedown and occupant 
restraint system, the dynamic responses (the restraint 
forces, the accelerations, and the excursions) of the 
wheelchair-occupant system imder crash loadings 
need to be analyzed. Computer simulations validated 
by sled tests have been used to study these dynamic 
responses. 

Wheelchair-Occupant Model 



the vehicle floor (or sled platform) are modeled as 
flexible planes to simulate the contacts between the 
dummy and the wheelchair, and the wheelchair and 
the veUcle floor. The dummy is modeled as a series 
of rigid segments and flexible joints as in standard 
ATB input files. Each rigid segment has 6 degree-of- 
freedom. The stiffness, damping (hysteresis and 
viscosity), permanent deformation, and friction of the 
contacts between the rigid segments and the flexible 
planes are determined by given contact functions. The 
functions can be of arbitrary form, linear or 
nonlinear. Figure 3 describes the model 
configuration. The static properties of the safety belt 
and tiedown (same webbing as the safety belt), were 
measured and shown in Figure 4. 

The sled test runs done at the UVA Automobile 
Safety Laboratory have been used to tune the model. 
The main purpose of this tuning process is to estimate 
unknown parameters such as the hysteresis and 
viscous damping properties of the contacts between 
the dummy and the wheelchair. 

Figure 5 shows the time histories of rear tiedown 
and shoulder belt forces. The level of agreement 
between simulation and sled tests is very good, 
particularly during the critical loading phase, 
considering that a comparatively simple model was 
used to describe a complex physical system. Errors 
in sled test data also contributed to the difference 
between simulation and sled test results. 



ERIC 



A computer model has been developed to simulate a 
wheelchair-occupant system tied down to a vehicle 
under the impact loadings of a crash. The surrogate 
wheelchair model has been developed to simulate a 
85kg/ 1871b powered wheelchair. The occupant model 
consisted of a 50th percentile Hybrid III dummy. The 
DYNAMAN (a version of ATB, Articulated Total 
Body Model) simulation packagefl] has been used to 
simulate the system. Figure 1 shows the surrogate 
wheelchair on a test sled. The computer model is 
illustrated in Fig. 2. In this model, the wheelchair 
mass is lun^>ed at its center of gravity with mass and 
mass moment of inertia around the three axes. The 
four wheels are modeled as rigid ellipsoids and 
connected to the wheelchair body by pin joints. The 
wheelchair cushion, back, armrests and footrests, and 

RESNA’94 • 



Model Validation 

The wheelchair-occupant and tiedown system was 
used for sled tests in following three laboratories: 
University of Virginia Automobile Safety Laboratory, 
University of Michigan Transportation Research 
Institute (UMTRI), and Middlesex University Road 
Safety Engineering Laboratory (MURSEL) (UK). 
The deceleration pulses of these three labs are shown 
in Fig. 6. The dynamic responses of the wheelchair- 
occupant system to these pulses are simulated using 
the model described here, and the results are 
compared with the test measurements. Table 1 shows 
the peak values of the restraint forces, accelerations 
and displacements of the wheelchair and dummy. The 
. correspondence between the simulations and sled tests 

75 

June 17-22, 1994 



61 



CRASH RESPONSE OF WHEELCHAIR 



is very good, although some values, such as the 
dummy chest vertical acceleration, for the MURSEL 
pulse show a significant discrq)ancy. Besides the 
modeling and measurement errors mentioned above, 
this discrepancy is also due to the poor repeatability 
of the MURSEL sled tests. 

Conclusions 

The wheelchair-occupant model developed in this 
paper can closely predict sled test results. It can be 
used to simulate crash responses^ and to estimate the 
effect of parameters such as crash pulse variation^ 
difference in sled platform rigidity^ and velocity 
change variations of sled pulses on output values such 
as the rear tiedown loads. 

References 

1. DYNAMAN User’s Manual, Version 3.0, Gesac, 
Inc. 1991 

2. Shaw, G, et al.., "Wheelchair Tiedown Inter- 
laboratory Test", Proceedings of the RESNA’ 94, 
June 17-22, 1994, Tennessee. 

Acknowledgments 

This research was sponsored by the National Institute 
on Disability and Rehabilitation Research (NIDRR). 





Pin joints 




Fig. 3 Wheelchair-Occupant nKxlel 




Fig. 4 Webbing force-elongation relation 



Fig. 1 Wheelchair surrogate on a sled 




BEST COPY AVAILABLE 






62 



RESNA ’94 



June 17-22, 1994 



CRASH RESPONSE OF WHEELCHAIR 



Table 1 Comparison of Simulation and Sled Test Results 



Peak Values of 
Parameters 


UVA 


UMTRl 


1 


VIURSEL 




Siinufai- 

Uoo 


Skd 

Tcm 


Difr. 

(%) 


Sfanub- 

Uon 


Sled 

T«t 


Diir. 

(%) 


Slimile- 

UOD 


8kd 

Ten 


Dur. 

(%) 


Left Rear Tiedown Force (lb) 


4620 


4594 


0.6 


4398 


4168 


5.5 


4109 


4238 


3.0 


Lap Belt Force (left) (lb) 


1789 


1730 


3.4 


1823 


1747 


4.4 


1660 


1746 


4.9 


Shoulder Belt Force (lb) 


2164 


2064 


4.8 


2107 


1885 


11.8 


2019 


1846 


9.4 


Wheelchair CO Acceleration (g) 


31.0 


30.7 


1.0 


31.0 


34.8 


10.9 


31.6 


33.0 


4.2 


Head ReaulUnt Acceleration (g) 


62.6 


56.3 


10.7 


63.5 


59.6 


3.0 


56.8 


53.7 


5.8 


Chest Vertical Acceleration (g) 


26.2 


23.0 


10.0 


31.0 


31.3 


11.8 


25.0 


40.2 


37.8 


Wheelchair CO Forward 


2.6 


3.0 


13.3 


2.6 


n/a 


n/a 


2.5 


n/a 


n/a 


Displacement (in) 




















Head CO Forward Displacement 


12.4 


13.6 


8.8 


11.3 


12.8 


11.7 


11.1 


n/a 


n/a 


Cm) 





















• All were obtained at the maximum forward moment with respect to the segment position at 



impact(at t=0). 



REAR TIEDOWNS 




SHOULDER BELT 




Fig. 5 Rear tiedown and shoulder belt forces. 
Experimental curves represent the average readings 



from three UVA sled tests. 



•■77 



UVA.UMTRI & MURSEL Pulses 




Time(s) 



— UVA 
UMTRl 

•• MUNBEI. 



Fig. 6 Sled deceleration pulses 



Walter D. Pilkey 
Morse Professor 
Transportation REC 
University of Virginia 
1011 Linden Av. 

Ouu'lottesville, VA 22903 U.S.A. 
Tele. (804) 924-3291 
Fax. (804)982-2037 



COPY AVAILABLE 



RESNA’94 • Jane 17-22, 1994 



63 



Wheelchair Tiedown Interlaboratory Test 



Greg Shaw, Arie Lapidot, Mike Seavnicky, John Thacker, University of Virginia; Larry Schneider, Miriam 
Manary, University of Michigan; Peter Roy, Geoff Savage, Chris Witherington, Middlesex University; 
Don Day, Defense and Civil Institute of Engineering Medicine 



ABSTRACT 

The deveiopment of standards for wheelchair 
tiedowns and occupant restraint systems for private 
vans depends on the definition of a dynamic sled test 
protocol that ensures repeatable results within a given 
laboratory and reproducible results among various 
laboratories, A multHab comparison study, involving 
frontal 30 mpb sled crash tests of a surrogate 
wheelchair and SOth percentile anthropomorphic 
dummy, was conducted in order to deteimine if the 
currently proposed test protocol was sufficiently 
defined. The results suggested that the test protocol 
achieved a level of precision comparable to that 
considered acceptable for automotive sled testing. 

INTRODUCTION 

Although there has been recent federal regulation 
regarding standards for wheelchair tiedowns and 
occupant restraint systems (WTORS) for transit and 
school buses (Americans with Disabilities Act transit 
regulations [56 FR 45530] and the National Highway 
Traffic Safety Administration (NHTSA) school bus 
guidelines), presently there is no standard for 
privately owned vans. The Society of Automotive 
Engineers (SAE) Adaptive Devices Committee 
Wheelchair Restraints Task Group is developing 
recommended practices which may lead to a national 
standard for vans, A key element of this task is the 
development of a dynamic sled test protocol that 
ensures repeatable results within a given laboratory 
and reproducible results among various laboratories. 
This will reduce the possibility that a WTORS which 
passes the prescribed test at one lab will fall the test 
at another, 

In its current form, the dynamic sled test protocol 
consists of three major components, the crash pulse, 
an acceleration / time curve that defines the severity 
of the crash, the surrogate (test) wheelchair that is 
representative of an 8S kg / 187 lb powered 
wheelehair, and the crash dummy that approximates 
a 67 kg / 163 lb male. Criteria for passing the test is 
defined by placing limits on the forward movement 
of the wheelehair and the dummy. 

Unlike the surrogate wheelehair and the crash 
dummy that have been narrowly defined, the crash 
pulse corridor was drawn to allow for a rather large 
variation in present crash sled deceleration 
characteristics. This raised concerns that the 



potentially large variation in the crash pulse might 
result in a similarly large variation in WTORS 
loading and that this variation may cause inconsistent 
results among laboratories. 

A multi-lab comparison test was conducted in 
order to determine if the proposed test protocol was 
sufficiently defined to produce reproducible results at 
different crash laboratories. 

MATERIALS AND METHODS 

The participating labs, locations for much of the 
WTORS testing conducted to date, included the 
University of Michigan Transportation Research 
Institute (UMTRI), Middlesex University (London), 
the Canadian Defense and Civil Institute of 
Engineering Medicine (DCIEM), and the University 
of Virginia (UVA). The pulse shape of each lab's 
sled, and the ability to tailor the shape, varied among 
the facilities. All sled carriages had nominally rigid 
steel or aluminum platforms and upper shoulder belt 
anchorage structures. 

Development of a Test Protocol 

In order to investigate whether parameters that 
were difficult to specify, such as the crash pulse 
shape, affected test precision, it was necessary to 
develop a test protocol that rigorously controlled all 
other test parameters. 

Test hardware 

UVA's version of the surrogate wheelchair, 
developed by Middlesex and the Transport Research 
Laboratory (TRL), was selected because it had 
proven sufficiently reliable, provided repeatable 
results, and approximated the weight of a powered 
mobility aid (85 kg) (fig. 1). In order to simulate the 
dynamics of commonly used 4-strap cargo tiedowns, 
a surrogate WTORS was designed using 7% 
elongation 48 mm wide polyester webbing. Both the 
surrogate wheelchair and the WTORS with integral 
tiedown load cells were shipped from lab to lab. Each 
lab was responsible for providing a SOth percentile 
Hybrid II test dummy that was in good repair. Each 
lab also provided load cells to measure occupant 
restraint belt tension and accelerometers to record 
sled carriage, wheelehair, and dummy decelerations. 

Crash specification 

,'I^^The collision was defined as it is in the current 



64 



RESNA -94 • June 11 ^ 2 , 2 , 1994 



Wheelchair Tiedown Interlaboratory Test 

SAE/ISO draft test procedures, namely a 48 +2, -0 
km/h frontal impact with the deceleration / time 
curve within the proposed sled crash pulse corridor 
(fig. 2). 

Test procedures 

A comprehensive set of instructions was 
developed for the participating laboratories. This 
package included assembly and installation 
instructions, instrumentation description and 
calibration information, data collection format, and a 
test protocol checklist. Hie test protocol was 
designed to carefully control the wheelchair and 
dummy position and WTORS pretension. 

Measured outputs 

In addition to the loads measured in the four 
tiedown legs and occupant restraint belts, 
accelerations were recorded for the sled, the 
wheelchair, and, multiaxially, for the dummy head 
and chest. Film analysis was used to determine the 
forward movement of the approximate chair P point, 
the intersection of the seat and the back, and the 
dummy head, hip, and knee. 




Figure 1- Test hardware Including (A) test dummy, 
(B) surrogate wheelchair, (C) surrogate strap 
tiedown, (D) Integral tiedown load cell, and (E) 
occupant restraint load cells. (F) locates the 
approximate wheelchair "P" point; (G) indicates the 
wheelchair center of gravity. 

RESULTS 

With some exceptions, the laboratories completed 
the test protocol without problems and without 
significant deviation, ftelimlnary analysis of 
DCIEM's forthcoming results indicated good 
agreement with the other labs' data. Table 1 
summarizes the test results In terms of peak recorded 
values. 

79 ' 

o 

ERIC qt r.nPY AVAILABLE RESNA'94 • 



DATA ANALYSIS 

The maximum values of all measured outputs 
were compared to determine the precision of the tost 
procedures. Precision is defined in terms of the 
repeatability, a measure of the variability between 
independent tests obtained within a single laboratory, 
and reproducibility, a measure of the variability 
between test results obtained at different labs. Both 
repeatability and reproducibility are commonly 
expressed In terms of the coefficient of variation 
(CV), which is defined as the standard deviation 
divided by the mean (table 1). Figures 2 and 3 
illustrate intra lab repeatability and inter lab 
reproducibility. 




Q 20 40 <0 so 100 120 

TIBIE(iiisec) 

Figured. The crash pulse corridor and the 
three UVA repeatability interlab crash pulses. 

A random effects statistical model was used to 
estimate the results of future tests conducted at 
another lab, The model uses within and between lab 
variance to produce a 95% confidence interval for 
results for fliture tests using the same inter lab test 
protocol (table 0. 



DISCUSSION 

Acceptable repeatability CV's for automotive 
safety testing usually do not exceed 1Q%- Standards 
fcr test dummy calibration place a maximum limit of 
10% for repeatability coefficients of variation (US 
Code of Federal Regulations, Vol, 49, Part S72). 
There Is general agreement that CV's of 1% are good 
[Versace 1983, Foster 1977], Most of the recorded 
outputs had repeatability CV's of less than S%; ell 
were less than 10%, This suggests that the precision 
of the Interleb tests for WTORS is comparable to that 
of state=of=the=art automotive sled testing, 

Because the overall repeatability and 
reproducibility was generally acceptable, the interlab 
protocol was generally successful in controlling the 



Jnne 1994 



65 



Wheelchair Tiedown Interlaboratory Test 



PARAMETER 


Interlab 

Average 


Repeatability 

CVV. 


Reproducibility 

CV% 


95% Lower 
Bound 


95% Upper 
Bound 


Left Rear Tiedown (IbO 


433S 


4.1 


6.2 


3735 


4934 


Shoulder Belt flbf) 


1932 


4.8 


7.1 


1624 


2239 


Left Lao Belt fibf) 


1741 


4.4 


5.1 


1657 


1825 


Sled Accel. (M 


24.1 


1.7 


4.6 


21.6 


26.6 


Head result, accel. (ifs) 


S6.S 


4.6 


6.5 


48.5 


64.6 


Chair x-accel. fa's) 


32.8 


3.4 


6.7 


27.8 


37.9 


Chair p-pt excursion (in) 


2.1 


9.5 


14 


1.3 


2.8 


Head excursion (in) 


13.7 


5.5 


12 


9.9 


17.3 


Hip excursion (in) 


6.0 


3.3 


12 


4.4 


7.6 



Table 1. Average peak values, repeatability and reproducibility CVs, and 95% predicted confidence 



intervals for fiituie test results. 




o ao 40 tfo 90 too tao 




TIME (p»««o) 

Figures. Interiab Reproducibility. Each curve 
is die average of die labs' 3 replicate tests. 



identified sources of variability. The three observed 
variations in sled crash pulse shape and in sled 
platform rigidity seemed to have little effect on the 
test results. This finding, along with the relatively 
limited predicted confidence intervals, suggests that a 
standard test protocol can be written that should 
minimize the possibility for a WTORS to pass the 
test at one facility while failing at another. 



REFERENCES 

Foster J, K.. Kortge J. 0., and Wolanin M. 1. "Hybrid 111- 
A Biomechanically-Based Crash Test Dummy," Safety 
Research and Development Lab, Environmental 
Activities Staff, General Motors Corp., SAE paper no, 
770938, 1977. 

Versace, J„ "The Motor Vehicle Manufacturers 
Association View on the Findings from NHTSA's Crash 
Test Repeatability Program," based on a Presentation 
Made on Behalf of The Motor Vehicle Manufacturers 
Association, NHTSA-Industry Meeting, Ann Arbor, 
Michigan, October 12, 1983. 

ACKNOWLEDGMENTS 

Members of the SAE Adaptive Devices Committee 
Wheelchair Restraints Task Group and the International 
Standards Association (ISO) WG-6 provided assistance in 
designing the test. Laurin Garland, TES Limited, 
coordinated the testing at DCIEM. John Aitchinson, UVA 
Department of Mathematics, developed the random effects 
model. This research was sponsored in part by the National 
Institute for Disability and Rehabilitation Research 
(NIDRR) and the Commonwealth of Virginia Center for 
Innovative Technology. 

Greg Shaw 
Transportation REC 
UVA Auto Safety Lab 
lOll Linden Av. 

Charlottesville, VA 22902 
804/ 296-7288 ph. 

804/ 296-3453 fax 



80 ; BEST COPY AVAILABLE 

^ ■ ■ 



66 



RESNA’94 • June 17-22, 1994 



GUIDELINES FOR WHEELCHAIR SECUREMENT AND PERSONAL RESTRAINT 



Thomas C. Adams, M.E., Steven I. Reger, Ph.D., Vinod Sahgal, M.D. 
Department of Physical Medicine and Rehabilitation 
The Cleveland Clinic Foundation, Cleveland, OH 



er|c 



Abstract 

Significant efforts have been made to improve the 
safety of wheelchair users while traveling in motor 
vehicles. There remain, however, many 
difficulties when the present wheelchair 
securement and personal restraint systems are used 
in public transit. A set of guidelines was 
formulated through consensus development of 
industry, research, and wheelchair using 
individuals that identifies the needs of all groups 
involved with the public transportation of 
individuals in wheelchairs. The guidelines address 
operation and performance characteristics of 
securement and restraint systems that will be 
acceptable and widely used. A test protocol to 
assure feasibility of the guidelines is also 
described. 



Background 

Much work has been done in recent years to 
evaluate the crash safety of wheelchair securement 
and personal restraint systems (1,2,3). These 
efforts provided a better understanding of how 
crash worthiness principles can be applied to 
wheelchair transportation to enhance the safety of 
all passengers. Standards and legislation have 
utilized the results from this work to identify 
aspects of wheelchair securement and personal 
restraint system that offer a high level of 
protection. 

The implementation of wheelchair securement and 
personal restraint technology on public transit 
vehicles, however, has encountered numerous 
operational difficulties that interfere with their 
proper use (4,5). Developing ’ acceptable 
securement and restraint systems for public transit 
is particularly difficult because the needs of many 
diverse groups must be considered. To address 
this problem, the Transportation Research Board 
has brought together these groups to reach a 
consensus on the heeds for securement and 
restraint systems that are acceptable for public 
transportation. 




Objective 

These guidelines provide outcome measures to be 
used for the development of wheelchair 
securement and personal restraint systems that are 
compatible with public transit. The guidelines will 
also be useful to transit systems to assist in the 
selection of securement and restraint systems. The 
validity of the guidelines will be demonstrated 
through testing of model securement and restraint 
systems. 

Approach 

The issues present in public tranportation that 
make successful wheelchair securement and 
personal restraint particularly difficult are listed in 
Table 1. 

Table 1 . Public transit issues 

Widely varying wheelchairs 
Widely varying individuals 
Varying vehicles designs 
Varying locations of wheelchair user 
Operator/wheelchair user interface 
Securement time 
Concerns of other passengers 

A resource panel was established consisting of 
recognized leaders from wheelchair users, transit 
providers, and manufacturers of wheelchairs, 
vehicles, and restraint systems. The diversity of 
the panel members assured that the needs of each 
group were considered in the guidelines. Each 
panel member provided expertise in a specialty 
area of securement and restraint. 

The guidelines that resulted from this effort 
provide outcome measures for performance and 
operation characteristics that are needed for 
acceptance in a public transit environment. The 
primary sections of the guidelines are given in 
Table 2. 

A model securement and restraint system will be 
developed and tested for conformance to the 
guidelines. The guidelines may be modified, with 



RESNA^94 • June 17-22, 1994 



67 



Wheelchair Secureraent Guidelines 



approval of the resource panel and the TCRP 
project panel , to assure that all the 
recommendations can be achieved in practice. 



Table 2. Securement and Restraint Guidelines 



Performance 


Operation 


Driving 


Information 


Impact 


Ergonomics 




Securement Time 




Feedback 




User Dignity 




Effect on Passengers 




Reliability 



Results 

Performance Guidelines 

The guidelines suggest that wheelchair securement 
and personal restraint systems should perform 
adequately xmder normal driving and impact 
conditions. The conditions to be evaluated are 
given in table 3. 

Specific test procedures are defined to evaluate the 
performance of securement and restraint systems. 
For each test, the wheelchair should be secured to 
a simulated vehicle floor, and a test dummy 
restrained in the wheelchair. For driving 
conditions testing, the guidelines suggest 
simulating the driving maneuvers through the use 
of either a tilt table or pull tester. Impact testing 
should be performed using a sled impact tester. 
Laboratory methods are recommended because 
they are more convenient, less expensive, and 
more reproducible than testing with a vehicle on a 
test course. 



Table 3. Performance Test Conditions 



Driving 


Impact 


Max. Straight Accel. 


20 g. Frontal 


Max. Straight Brake 


5 g. Rear 


Double Lane Change 


5 g. lateral 


Right Turn 




Left Turn 





The parameters to be evaluated during the testing 
are given in table 4. Both tests measure the 
response of the test dummy to predict the lesponse 
of an individual in a wheelchair. The driving test 




evaluates comfort and the impact test evaluates the 
potential for injury. 



Table 4. Performance Test Parameters 



Driving 


Impact 


Wheelchair Displ. 


Wheelchair Displ. 


Test Dummy Displ. 


Test Dummy Displ. 


Test Dununy Accel. 


Test Dummy Injury 


Wheelchair Tipping 




Dummy /Bus Contact 





Operation Guidelines 

The section of the guidelines that looks at 
operation issues deals with the aspects of 
securement and restraint systems that enable them 
to be accepted for use in a public transit 
environment. It is important that the individual in 
the wheelchair and any individual who may be 
assisting with the securement or restraint process, 
be able to use the system correctly and quickly. 

Adequate information is the first step toward the 
efficient use of a securement and restraint system. 
The guidelines recommend that information be 
provided through brochures, manuals, training 
sessions, and on-board instructions. 

The securement and restraint system should 
operate easily to reduce securement time, prevent 
occupational injury, and encourage user 
independence. The guidelines identify easily 
accessible locations for system components that 
require handling, and maximum forces to operate 
the system. 

For the safety of all passengers on the vehicle, the 
wheelchair must be adequately secuied. The 
securement system should have a feedback 
mechanism to notify the vehicle operator when the 
wheelchair securement is complete. 

The guidelines recommend that a securement 
system should be able to be attached to most 
wheelchairs within sixty seconds. They also 
recognize that the securement system should be 
able to be released in less than fifteen seconds in 
an emergency. A redundant release should be 
included to release the wheelchair from the 
securement system if the normal release 
mechanism malfunctions. 



68 



RESNA ’94 • June 17-22, 1994 



Wheelchair Securement Guidelines 



The guidelines further suggest that securement and 
restraint systems be designed to minimize the 
contact between an individual who is assisting 
with the securement and restraint and the 
individual in the wheelchair. It is indicated that 
this can be encouraged by locating the parts that 
require handling in the ergonomic locations 
identified in the guidelines. 

It is important that the activities of the other 
passengers be minimally affected by the presence 
and use of a securement and restraint system. The 
guidelines specifically recommend attention to 
preventing hazards and interference with the 
seating or movement of other individuals on the 
vehicle. 

Well maintained and functioning securement and 
restraint systems are vitally important to the safety 
of all individuals on the vehicle. The reliability of 
these systems is addressed in terms of 
maintenance, tamper resistance, and durability 
against aging, wear, and the elements. 

Discussion 

When evaluating the performance of securement 
and restraint systems, it is desired to predict the 
response of the wheelchair users to indicate a level 
of comfort and injury protection. The approach 
suggested here is to identify threshold levels of 
motion and forces of key anatomical locations that 
can be clinically associated with comfort and 
injury tolerance. Instrumented test dummies then 
provide test data for comparison to these threshold 
levels. 

The ability to accurately predict motions and 
forces applied to wheelchair users based on the 
response of test dummies, however, is a significant 
concern. The use of test dummies is suggested 
here to provide a reproducible test methodology 
that can be used to compare the performance of 
securement and restraint systems. The correlation 
between the test dummy response and that of 
wheelchair users is an area that needs additional 
research. 

There is little data available on the real life success 
of the current securement and restraint systems in 
life-threatening accidents. Test data indicates that 
if the systems are applied correctly, they will be 
highly effective. A securement and restraint 
system is only effective protection, however, when 



it is used correctly. All of the factors presented 
here must be considered when systems are 
designed and selected so that their correct use 
does not conflict with the other demands in a 
public transit environment. 

Many of the issues that are addressed in these 
guidelines are not identified in other standards, 
guidelines, and legislation, but they are critical 
issues to consider if wheelchair travelers are to be 
efficiently accommodated in public transit. 

References 

1. ECRI, Positioning and Securing Riders with 
Disabilities and Their Mobility Aids in Transit 
Vehicles: Designing and Evaluation Program - 
Final Report, Project ACTION, Washington, DC, 
Nov. 1993 

2. Schneider LW, Rationale, Historical Synopsis 
of the Literature, and Bibliography, Vol. 2., 
Biokinetics and Associates, Ontario, Dec. 1991 

3. Dalrymple G., et. al.. Wheelchair and Occupant 
Restraint on School Buses, Report No. DOT-TSC- 
NHTSA-90-1, National Highway Traffic Safety 
Administration, US Department of Transportation, 
Washington, DC, 1990 

4. Turner M, Disabled Bus Riders Protest New 
Rules, The Davis Enterprise^ Sacramento, CA, 
August 9, 1993 

5. Reger S, Adams T, Comparative Field Testing 
of the Cleveland Securement System, Project 
ACTION Report y Washington, DC, 1993 

Acknowledgments 

The authors wish to acknowledge project funding 
through TCRP Project C-1, administered by the 
Transportation Research Board. 

Thomas C. Adams 

Dept, of Physical Medicine and Rehabilitation 
Cleveland Clinic Foundation 
9500 Euclid Ave. 

Cleveland, OH 44195 



83 



RESNA ’94 • June 17-22, 1994 



69 



CAR ADAPTATIONS FOR PEOPLE WITH SPECIAL NEEDS IN EUROPE 
A SIMULATOR STUDY EVALUATION 

Roelof Veenbaas, TNO Road Vehicles Research Institute, Delft 
Willem B. Verwey, TNO Institute for Human Factors, Soesterberg 
the Netherlands 



ABSTRACT 

Adapted cars are used to enhance the independent 
mobility of People with Special Needs (PSN). ATT 
systems are designed to increase comfort and safely of 
able-bodied drivers. The requirements of Drivers with 
Special Needs (DSN) are mostly not included in the 
design of these systems. The influence of ATT systems 
on workload and driving performance of DSN in 
adapted cars should be tested as well. Driving simulator 
experiments were performed to obtain a baseline for 
testing ATT systems in future. 



BACKGROUND 

Independent mobility is an important factor which 
influences the quality of life of People with Special 
Needs (PSN). One way to improve independent 
mobility for people with impairments is the use of 
adapted cars. More and more Advanced Transport 
Telematic (ATT) systems, like automatic distance 
keeping and intelligent cruise control, are developed to 
make driving a standard car comfortable and safe. 
Basic MMI (Man Machine Interaction) design of new 
cars is often based solely on the requirements of 
able-bodied drivers, while DSN (Drivers with Special 
Needs) are normally not considered. Adaptation 
requirements and the technical realization relies on 
skilled craftmanship, to suit individual DSN with 
specific needs. A gap exists between the resources and 
market size of standard car industries and of the 
car-adaptation industries. In an effort to bridge this gap 
the aim of the TELAID (Telematic Applications for the 
Integration of Drivers with Special Needs) project has 
been to describe the requirements of Drivers with 
Special Needs (DSN) on the one hand (1) and 
investigate the usability of telematic applications in 
facilitating driving for DSN on the other hand. 
Integration of telematic applications could decrease the 
drivers workload and facilitate driving by PSN who are 
not able to drive yet. Design guidelines for the MMI 
interface of telematic applications, taking into account 
the requirements of DSN, are essential. 

The aim of the first phase simulator experiments, 
described here, was to find the baseline performance of 
car-adaptations currently used by DSN. Driving 
performance has been subject to research before. Yet 
little research has been done to investigate the influence 




RESNA ’94 



of car-adaptations on the driving performance and 
workload of DSN. 

RESEARCH QUESTIONS 
The intention of this study was to analyze wether 
specific control aids conform to the specific needs of a 
group of DSN and not to assess wether certain 
impairment groups should be restricted or not be 
permitted to drive. The main research question was to 
which extend the use of car-adaptations, especially 
hand controls, influences the performance and 
workload of DSN. 

The following hypotheses were formulated: 

1. It is expected that there is a difference between 
DSN using a hand controlled brake and 
accelerator and able bodied drivers using pedal 
control in the reaction to emergency situations 
(brake reaction time). 

2. It is expected that the general (overall) driving 
performance is the same for the able-bodied 
drivers and DSN. For certain components of the 
driving task differences are expected. 

3. It is expected that DSN using full manual control 
experience a higher mental and physical workload 
compared to able-bodied drivers using 
conventional pedals. 

4. It is expected that there are differences in the 
reaction to emergency situations, the performance 
of the driving task and mental and physical 
workload due to the type of full manual control 
used. 

The simulator experiments were performed at the 
Swedish Road and Traffic Research Laboratories (VTI) 

(2) and at the TNO Institute of Human Sciences, (IZF) 

(3) in the Netherlands. The chosen research methods 
were similar. At VTI a moving base driving simulator 
was used and at TNO/IZF a fixed base driving 
simulator. Method and results from the Dutch 
experiments will be discussed here. 

METHOD 

The design of the experiments was a classical single 
factor design with an experimental and a control group. 

Subjects 

Testing all categories of drivers with special needs was 
not possible within the time and resource limits of the 
project. Subjects were selected from the largest category 



70 



June 17-22, 1994 



EVALUATING CAR ADAPTIONS FOR PSN 



of DSN using car adaptations, people with a lower limb 
impairment. From the requirements inventory of people 
with special needs (1) it became clear that the visual 
impairment group as well as the lower and upper 
im p airme nt groups presented relatively large problem 
areas. The vision capabilities of car-simulators are 
somewhat limited, which is another indication for the 
choice of mobility impaired subjects. The IZF target 
group consisted of experienced paraplegic drivers. The 
manual controls used were similar to the ones the DSN 
use in daily practice. 

Thirty subjects participated in the Dutch study. Ten 
subjects were used to drive with a segmented grip hand- 
controlled accelerator under the steering wheel, ten 
subjects used a ring accelerator above the steering 
wheel. Both experimental groups used the same hand 
controlled brake. 




The control group consisted of able bodied drivers 
using pedal control of accelerator and brake. 



Driving task 

A three level - strategic, manoeuvre and control - 
driving task definition (4) was extended for DSN, to 
• include additional tasks like access (entering and 
leaving the car) and car maintenance (1). The driving 
task for the simulator study was composed of 
components from control (steering, braking etc.) and 
manoeuvre (trafBc interaction) level (5). 

The experimental road consisted of a two lane 80 km/h 
road of 80 km long, 7.2 m wide, which is common in 
the Netherlands. It included sections with many curves 
and sections with few curves with oncoming cars and 
cars parked in the driving direction. Some of the parked 
cars started to merge. Red and yellow squares 
appearing just above the horizon were used to test 
emergency braking. Subjects were instructed to brake as 
fest as possible to the red squares and ignore the yellow 
ones. 

All subjects performed the same driving task. A 
practice route that was driven before the experimental 
route had similar characteristics, but no merging cars. 



Measurement methods 

Driving performance was evaluated comparing brake 
reaction times and comparing the mean and variation of 
the speed and of the lateral position of the road under 
six different driving situations. 

Physical Discomfort in the upper body area, especially 
the hands was measured using an adapted version of the 
LMD scale introduced by Van der Grinten (6). On the 
LMD scale subjects indicated a physical discomfort 
level before and after the experiment. 

Subjects had to rate workload aspects, mental demand, 
physical demand, time pressure, performance, effort 
and frustration, using the Raw Task Load Index 
(RTLX) (7). The RTLX was filled out for the driving 
task as a whole and for six more specific situations. 
Subjects filled out two questionaires, one before the 
experiment on driving experience and age and one after 
the experiment on the comparability of the simulator 
driving to driving their own car. 

Since most DSN came to the simulator in their own car 
it was possible to compare the control forces required 
for accelerating and braking in the simulator to the 
forces required for accelerating and braking their own 
car. 



RESULTS 

Minor differences, significance (p<0.05), were found in 
the driving performance : The ring accelerator group 
showed a larger standard deviation of speed during 
straight sections and the driving of sharp curves and 
also a larger standard deviation at the steering wheel 
position and increased lu* of braking actions during the 
overtaking of parked cars. 

The LMD scale showed a clear physical discomfort 
increase for the group using the segment accelerator in 
the fingers of the right hand. 

No differences in workload (RTLX) were found 
between the groups. 

An overview of the control forces [N] for both 





group 


accelerator 


brake 


minimal 

stroke 


full 

stroke 


minimal 

stroke 


full 

stroke 


mock 


segment 


10 


20 


20 


45 


up 


ringac. 


10 


25 


20 


45 


own 


segment 


23 ±12 


45 ±17 


14±7 


40± 14 


car 


ring ac. 


19± 12 


39 ±13 


18± 11 


37± 15 



Although there is no absolute norm for safe driving, 
from an overall view of driving performance in the 
simulator tests, we subjectively estimate that differences 
found in this study are within the safety limits. 



8i 



RESNA ’94 



June 17-22, 1994 



71 



EVALUATING CAR ADAPTIONS FOR PSN 



DISCUSSION 

The experiment was carried out to evaluate the merits 
of conunonly used adaptations for paraplegic drivers. 
The method presented here has proven to be useful. 
Experienced DSN (mobility impaired), with well 
designed adaptations for the primary control of their car 
perform very much the same as able-bodied drivers 
driving standard cars. Though there seem to be 
indications that they do it at a higher cost. Differences 
in performance could be partly traced back to the way 
the car was adapted with the ring accelerator. A proper 
design and installation of hand controls combined with 
a car with low level pedal forces can contribute to 
decrease the level of discomfort and fatigue on long 
driving distances. 

The reported study forms a necessary baseline for 
further simulator studies regarding driving performance 
of DSN. Introduction of Advanced Transport 
Telematics in private cars can in a crucial way enhance 
mobility for DSN but it might as well ruin what has 
already been gained. This would be the case if the ATT 
systems do not correspond to the needs and abilities of 
people with special needs. Adaptation for DSN is often 
based on ad hoc principles and very seldom any 
evaluation of the system (driver + adapted car) is 
carried out. The use of adequate and common 
assesment methods, design guidelines and legislation 
could facilitate an optimal use of ATT systems and 
establish safety requirements. A natural step on the way 
to develop assessment methods would be to continue the 
simulator studies by introducing ATT systems. The 
selection of these systems could be based on two 
principles: 

1 . How could ATT systems enhance driving 
performance or comfort for DSN, 

2. How do different interaction modalities due to the 
introduction of ATT influence driving 
performance of DSN. 

Finding a way to incorporate DSN requirements into 
the design of standard ATT components could resolve 
the indicated gap and should be strived for. This 
principle is valid for many rehabilitation fields as was 
mentioned before by Vanderheiden (8). 

REFERENCES 

1. Nicolle C., Ross T., Richardson S. (eds.). 
Identification and Grouping of Requirements for 
Drivers with Special Needs. CEC, DRIVE II 
Project V2032 TELAID, deliverable 3, 1992. 

2. Peters, B., Nilsson L., Driving Performance of 
DSN using hand controls for braking and 
accelerating. Proc. of the 26th ISATA symposium, 
Aachen, Germany, September 1993. 



3. Verwey, W.B. & Veenbaas, R, Driving 
performance of paraplegic DSN using hanH 
controls for braking and accelerating, in: R 
Veenbaas (ed.) Validation of the identification of 
special needs of DSN. CEC, DRIVE II Project 
V2032 TELAID, deliverable 5, 1993. 

4. Michon, J.A., A critical view of driver behaviour 
models: what do we know. What should we do. In 
L. Evans and RC. Schwing (eds.). Human 
behaviour and traffic safety (pp 485-520) New 
York: Plenum, 1985. 

5. Verwey, W.B,, On evaluating vehicle adaptations 
for disabled drivers, in: R. Veenbaas (ed.) 
Validation of the identification of special needs of 
DSN. CEC, DRIVE II Project V2032 TELAID, 
deliverable 5, 1993. 

6. Van der Grinten, M.P, & Smitt P., Development 
of a practical method for measuring body part 
discomfort. In S. Kumar (Ed.), Advances in 
Industrial Ergonomics and safety IV. (pp. 

3 1 1-3 17). London: Taylor & Francis, 1992. 

7. Byers, J.C., Bittner Jr, A.C & Hill, S.G., 
Traditional and raw task load index (TLX) 
correlations: Are paired comparisons necessary? 

In A. Mital (Ed.) Advances in Industrial 
ergonomics and Safety I (pp. 481-485). London: 
Taylor and Francis, 1989. 

8. Vanderheiden, G,C,, Keynote address. Ecart 2 
conference, Stockholm, Sweden, 1993. 

ACKNOWLEDGEMENT 
Funding of the TELAID project is provided by the 
Commission of the European Conununities DG XIII, 
within the DRIVE II programme. The TNG Road 
vehicles research Institute contribution is coftmded by 
the Dutch Ministeiy of Social Affairs and Employment. 

The authors wish to thank the TELAID consortium for 
their contribution to this study, especially BjOm Peters 
(Swedish Road and Traffic Research Laboratories - 
VTI, Sweden) for his contribution to the development, 
execution and evaluation of the simulator experiments 
and Colette Nicolle (The HUSAT Research Institute, 

UK) and Manfred Dangelmaier (University of 
Stuttgart/IAT, Germany) for their contribution to the 
evaluation of the experiments. 

Roelof Veenbaas, 

TNO Road-Vehicles Research Institute, 

Schoemakerstraat 97 

P.O. Box 6033, 2600 JA Delft, the Netherlands 

Fax: +31 15 6973 14 

Email: Veenbaas@IW.TNO.NL 



72 



RESNA ’94 



' 86 

• June 17-22, 1994 



A SMALL SCALE DRIVER EVALUATION VEHICLE 



David H. Pearce, Ph,D., P. E., E. H. Harrison, Jr., Ph.D., Bronwyn Keller, OTR/L. 
Mississippi Methodist Rehabilitation Center 
Jackson, Mississippi 



Abstract 

To provide an intermediate level driver evaluation 
vehicle, a golf cart based small scale vehicle has 
been built. This Small Scale Vehicle (SSV) 
contains extensive safety features to protect the 
client and the instructor. The instructor can obtain 
full control of the vehicle at any time. Broad 
testing capabilities verify the client’s performance. 
The design required a vehicle for use with stroke, 
spinal cord, and traumatic brain injury patients. 
An adaptive steering wheel, hand controls, and left 
accelerator pedal allow a wide range of patients to 
use this device. The equipment has been built and 
is in use. Currently we are gathering data on 
patient and equipment performance to use as a data 
base for future patient comparisons. 



Problem Definition 

As patients adapt to selected injuries, there is often 
a need to evaluate their driving skills, with training 
or retraining as appropriate. Because of the wide 
variability and unknown responsiveness of patients 
to highway driving, it was decided that a safer 
environment was necessary than a highway activity 
provides. Certain classes of patients, such as the 
brain injured, might exhibit acceptable behavior in 
the classroom, but not be able to perform in a 
moving vehicle. Some kind of actual driving was 
deemed necessary, yet controlled conditions were 
equally important. 

In an effort to provide this opportunity, a small 
scale driver evaluation vehicle was developed. 
Aside from allowing basic driver evaluation, 
modifications make measurement of certain 
responses under safe driving conditions possible. 
Safety of the patient was of utmost concern in the 
design of the vehicle. Also, since vehicles with 
zero effort systems are nearby' it was decided to 
concentrate on a vehicle which would meet the 
needs of most of our clients. The desired vehicle 
would also fill the gap between our road vehicles 
and the classroom based testing equipment. It 
would be adaptable to a wide range of tests, adapt 
to different types of physical needs, and drive over 



grass. 



Rational Design 

Safety and functional testing guided the vehicle 
design. As a platform, the Club Car golf cart 
powered by a 36 volt DC motor was chosen. The 
safety goils included bucket seats for patient 
support, automotive style lap and shoulder belts, 
amber flashing light on roof, electrical interlock so 
that vehicle could not be started without an 
instructor, windshield for protection, rear view 
mirrors for training and evaluation, steel cage 
around occupants, headlight, brake lights, and turn 
signals. 

The functional evaluation equipment included 
adjustable hand controls (braking and acceleration), 
adjustable position steering wheel with an 
assortment of hand attachments, a left foot 
accelerator, an instructor’s brake, hand parking 
brake, reaction timer with various conditions, 
instructor timer control, random timer control, 
instructor selected stop light pattern, left turn 
avoidance test, manual control of stop lights, and a 
remotely operated stoplight. The instructor 
controlled events can be initiated without informing 
the client. 

Development 

The items in the rational design were integrated into 
the structure of the basic golf cart to produce the 
functional unit illustrated below. Instructor 
controlled parameters are divided between a control 
panel on the right front of the unit and an instructor 
held pendant. The control panel is shown below. 
The upper left of the panel contains the main power 
controls, amber light switch, and a circuit breaker 
to protect the instrumentation. The digital timer 
readout is in the upper tight. The lower half of the 
panel contains switches which set up the test 
conditions. One switch determines whether the 
instructor can delay this event until the driving 
conditions are appropriate. The next switch 
determines whether a combination of the four cart 
mounted stop lights are used or whether a remote 

^7 



RESNA ’94 • June 17-22, 1994 



73 



SMALL SCALE VEHICLE 




stoplight is used. There is a timer enable switch. 
There are two different tests for response of the 
patient. The patient can respond by stepping on 
the brake or by executing a sharp left turn 
avoidance manoeuver. The Test Brake/Tum 
Switch sets up these conditions. A si gnal 
Manual/Timer Switch selects either the timer or 
the operator initiated activity from the hidden 
hand-held pendant. A timer reset button permits 
resetting the timer after use. The seat safety 
switch can be set to require a person in the right 
seat before the vehicle will move. This will 
prevent an over-eager student from making the 
vehicle move before the instructor is in place. 
There is also a safety switch in the driver’s seat so 
that the vehicle can not be made to move if there 
isn’t a person in this seat. 

The pendant was designed so that it was quiet in 
operation and could be held on the right side of the 
instructor out of the client’s view. The pendant 



contains four lever switches and one push button 
switch. The push button switch activates the si gnal 
lights in the manual mode and the timer sequencer 
in the timer mode. 

Four red signal lights are located on the SSV. 
Each 1 1 mm diameter light is mounted in a 5.5 cm 
by 8.4 cm by 3.8 cm box which has two suction 
cups attached to it. The selection of these lights is 
controlled by the instructors pendant. The suction 
cups allow positioning of the lights at various 
locations on the windshield and on a clear plastic 
piece across the back of the vehicle. The rear 
lights can be positioned to be in the patient’s field- 
of-vision when looking in the rear-view mirror. 
The lights can be used to evaluate the patient’s 
visual perceptual ability under actual conditions. 

The hand controls are Supergrade IV by Handicaps, 
Inc. A left foot accelerator is also from Handicaps, 
Inc. An assortment of quick release spinners by 
Greshman Driving Aids, Inc. is available for the 
client’s use. The adjustable steering column was 
manufectured to our specifications by Drive Master. 
The electronic timer is the standard AAA 
Automatic Brake Reaction Timer. This timer was 
modifted to work on 12 volt DC. 

Evaluation 

The vehicle has been in use for over a year. Over 
25 spinal cord, stroke, and traumatic brain injury 
clients have used the vehicle in various types of 
training activities. The adaptability has been 
verified by the variety of patients who have used 
the vehicle. An instruction manual has been written 
for reference by the staff. It takes approximately 
one hour of instruction for a therapist skilled in 



ftCSET POWCR 

# m 


TOP ♦ 

LIGHT 

# 


SIGNAL 








TIME 


1 

SEAT W 




TIMER 


TIMER 


TEST 


SIGNAL V 


TIMER 


SAFETY M 


% 


MANUAL 


LOCAL 


ON 


BRAKE 


MANUAL 


RESET 


ON 1 




RANDOM 


c m 

REMOTE 




Or 

TURN 


d 

TIMER 


Q 


^ } 



BEST COPY AVAILABLE 

® 74 

ERIC 




EESNA ’94 • June 17-22, 1994 



SMALL SCALE VEHICLE 



driver training to be able to use the vehicle to 
evaluate a patient. One class of patients that this 
vehicle is particularly suited to is the young spinal 
cord Injury patient who has never driven. This 
vehicle provides a safe environment in which basic 
skills can be learned. Patients and therapists like 
the ease of entry, the bucket seats, safety belts, 
and the general feeling of security provided by a 
small, slow, vehicle which is operated off of the 
roadway. 



Discussion 

The development and use of a small scale driver 
evaluation vehicle has been presented. This 

vehicle meets the needs of our patients at an 
appropriate technology level. The SSV fills an 
important niche in the range of available devices. 
Clients feel safe in this device and are not afraid 
and apprehensive as they might be in a fiilLsized 
vehicle. The excellent visibility and feeling of 
safety make the SSV fun to drive. The 

characteristics of the SSV permit tests which could 
not be safely performed with a fiill-sized vehicle. 
The SSV also gives the instructor the oppormnity 
to experiment with different control settings in a 
safe environment. 

References 

1. Hale, P.N., J.R. Schweitzer and M. Shipp. 
(1987). A Small Scale Vehicle for Assessing and 
Training Driving Skills Among the Disabled. 
Arch Phvs Med Rehabil . Vol 68, 741-742. 

Acknowledgement 

This work is supported by internal funds of the 
Mississippi Methodist Hospital and Rehabilitation 
Center. 

Address 

David H. Pearce, Mississippi Methodist 
Rehabilitation Center, 1350 East Woodrow Wilson 
Dr., Jackson, MS 39216. 

(601) 364-3371 








RESNA ’94 



June 17-22, 1994 



75 



A HUMAN POWERED VEHICLE FOR THE PH\'SICALLY DISABLED 



John M. Henshaw and Stefan E. Radloff 
Department of Mechanical Engineering 
The University of Tulsa, Tulsa, OK 



ABSTRACT 

A unique, three wheeled, human powered vehicle 
(HPV) was designed and constructed for an eight 
year old boy with spina bifida. The design allows 
the boy, Kramer, to propel the tricycle using a 
reciprocating motion. This simple motion is one 
that Kramer can accomplish, in contrast to the 
more complex rotating motion used in 
conventional tricycles and bicycles. 

The tricycle also has specially designed pedals to 
hold Kramer's feet. A five speed internally geared 
hub provides a low gear for climbing hills as well 
as high gears for higher speeds. This HPV is the 
first of several tricycles that Kramer has been able 
to ride, providing him with fun, exercise, 
responsibility, and independence. 

BACKGROUND 

The recipient of the tricycle, Kramer, is an active 
eight-year-old boy with spina bifida. Kramer has 
no feeling in his feet, and he wears braces to keep 
his feet aligned forward. Kramer can walk and run 
for short distances, although he tires easily, since 
some of the muscle groups in his legs are 
underdeveloped, especially the hamstring muscles 
on the back of the thigh. As a result, Kramer has 
difficulty keeping up with other children his age. 
Also, despite repeated attempts, Kramer appears to 
lack the coordination to perform the rotating leg 
motion required to propel a conventional bicycle or 
tricycle. 

Kramer's parents have previously purchased a 
number of tricycles, including a very expensive 
device custom-built for Kramer. In all cases these 
machines have proven to be ineffective, as they all 
had pedals that required rotating motion. Also, 
Kramer had difficulties keeping his feet positioned 
on the pedals. Kramer's attempts at fun and 
exercise fi-equently ended in fhistration as a result 
of these problems. 

After meetings with Kramer and his parents, and 
observing Kramer attempt to ride his older tricycle, 
several design objectives became apparent. The 



new HPV would require a totally different pedaling 
mechanism so that Kramer could propel the 
vehicle; one that did not depend on rotating 
motion. The design team's testing of Kramer 
showed he is particularly strong in leg extension, 
especially when he can push against a back 
support. The drive mechanism for the prototype 
should take advantage of this. Foot supports to 
keep Kramer's feet on the pedals were also 
important. 

’ STATEMENT OF THE PROBLEM 

The freedom and responsibility offered by a tricycle 
or bicycle is an important part of the development 
of many children. The design team's goal was to 
design, construct, and evaluate a device that would 
not only provide Kramer with exercise and 
habilitation, but also freedom and responsibility. 

As discussed above, foot support and the use of leg 
extension were significant concerns. Several other 
important overall goals were also considered 
throughout the design process. Safety was of 
course a primary concern. Also, the vehicle was to 
be as lightweight and portable as possible. 

Another significant concern was to ensure that 
Kramer would not quickly outgrow the tricycle. 
Thus, adjustability of the seat and pedals were 
addressed. When possible, off the shelf parts were 
used to help minimize repair and machining costs. 
Finally, it was important to make the vehicle fun. 
For Kramer, fun was synonymous with fast. The 
designers tried hard to make the vehicle look 
"cool", like a different bicycle instead of a 
rehabilitation device for the disabled. 

DESIGN AND DEVELOPMENT 

A team of five senior mechanical engineering 
students designed and built the HPV according to 
the objectives described above. Detailed results are 
contained in [1]. 

After consideration of numerous mechanism 
concepts, a design using reciprocating pedals was 
chosen for use in the HPV. The final concept is 
shown schematically in Figure 1. 




• June 17-22, 1994 



76 



RESNA ’94 



Human Powered Vehicle 



Force is applied to the pedal and transmitted down 
the crank arm to the cable crank. The force is then 
transmitted along the cable and into the one way 
clutch mechanism. The clutch locks in the forward 
direction, and free-wheels in the reverse direction. 
When pressure is released from the pedal, a return 
spring pulls the pedal back to its forward position. 
A schematic showing the overall drive mechanism 
is shown in Figure 2. 



pedal 




Figure 1 - Schematic of the pedaling mechanism 

The force exerted on the pedals is transferred into 
two one way clutches as previously described. 
Force is then transferred via a chain into the five- 
speed internally geared hub. This hub features a 
single lever, indexed shifting with the shifter 
located conveniently on the handle bars of the 
tricycle. From the five speed hub, force is 
transmitted through the drive chain to the rear axle 
by means of a keyed gear. An important feature of 
the mechanism is that the two pedals are not linked 
together. One pedal can be pushed down while the 
other is up, or both can be pushed at the same 
time. 




Figure 2 - Overall schematic of the drive 
mechanism 



Using the mechanism described, a preliminary test 
model was constructed by modifying an industrial 
grade (adult’s) tricycle. The test model proved, 
most importantly, that the proposed mechanism 
worked. Kramer could not only ride the test 
model, but thoroughly enjoyed riding it, even 
though it was extremely heavy and ungainly. The 
team observed Kramer extensively while he rode 
the test model, both on a parking lot and indoors 
on a modified set of bicycle "rollers". These 
observations were critical in modifying the design 
for the final prototype tricycle that was delivered to 
Kramer. From the test model, it was determined 
that Kramer could exert a greater force on the pedal 
at the end of the stroke than at the beginning. 

Thus, the configuration of the cable crank was 
changed firom a constant radius to an increasing 
radius, giving Kramer a greater mechanical 
advantage (higher torque) at the beginning of the 
stroke. 

Also, the ideal design, placement, and travel of the 
pedals were determined using the test model. The 
pedals themselves were developed not only to 
secure Kramer's unfeeling feet against slipping, but 
also to allow him to get his feet in and out of the 
pedals quickly and safely without assistance. The 
pedals consist of off-the-shelf bicycle pedals with 
inexpensive molded plastic guides attached to them 
to accomplish the above goals (the guides are not 
shown in Figures 1 and 2). Seat position, location 
of the handlebars, crank arm length, pedal 
placement, and head (steering) tube location were 
also established using the test model. 

With the information and measurements gathered 
from the test model, a prototype tricycle was 
manufactured. The prototype can be described as a 
recumbent tricycle with a reciprocating crank 
mechanism. The frame was constructed from 1.25 
in. outside diameter 4130 Cr-Mo steel tubing. The 
tricycle features 20 in. aluminum alloy rims on all 
three wheels, a five speed internally geared hub, 
and a single drum bi^e on the front wheel. The 
completed prototype tricycle is shown in Figure 3. 

EVALUATION 

Much of the testing occurred during the design and 
development phase of the project with the test 
model. The prototype tricycle was initially tested 
with Kramer riding it in an indoor hallway. The 
tricycle worked well, and was shortly thereafter 
delivered to Kramer for use at home. Kramer 
instinctively pushes the pedals with both feet 
together when starting from a stop. This gives 
way to alternate-foot pedaling as he speeds up. (He 
was not trained in this pedaling technique - it just 
seems to happen naturally.) As with most 



91 

RESNA’94 • June 17-22, 1994 



77 



Human Powered Vehicle 



children, steering and braking required a little 
practice for Kramer to master. Early on, he "field 
tested" his tricycle with several harmless collisions 
with ctitbs and other stationary objects. 




Figure 3 - The completed prototype tricycle 



Kramer has since subjected the tricycle to many 
miles, including several family outings of more 
than three miles of riding! To date no si gnifican t 
problems have developed with the prototype. 
Several minor adjustments of the five sp^ hub 
have been made, otherwise the tricycle has 
performed very well. 

DISCUSSION 

The prototype tricycle successfully meets the major 
goals of the project described earlier. About $1800 
was spent on parts, materials, and some of the 
machining. (The above figure does not include 
engineering time or the bulk of the machine work, 
which was donated.) 



machine that is adaptable to as many children as 
possible while minimizing the cost of the machine. 

REFERENCES 

1. Radios', S., Klein, G., McClendon, C., Smith, 
L., and Swank, S., "A Human Powered Vehicle for 
the Physically Disabled", The University of Tulsa 
Department of Mechanical Engineering Project 
Report, April, 1993. 

ACKNOWLEDGMENT 

Fimding for the HPV was provided by the Tamara 
Lilly Brown Memorial Fimd of the University of 
Tulsa. Additional funding was provided by the 
University ofTulsa Office of Research. McElroy 
Manufacturing, Inc. of Tulsa, Oklahoma donated 
fabrication services. Kramer and his parents 
provided the cooperation and enthusiasm that made 
this project exciting and rewarding for the entire 
project team. 



John M. Henshaw, P.E., Ph.D. 
Assistant Professor 

Department of Mechanical Engineering 
University of Tulsa 
600 South College Ave. 

Tulsa, OK 74104 

(918)631-3002 

me jmh@vaxl.utulsa.edu 



A second generation tricycle is currently being 
designed for a young cerebral palsy patient by a 
new team of mechanical engineering students. 
While the current model satisfied all of the design 
requirements, several improvements are 
contemplated. At about 50 poimds, the tricycle 
weighed more than expected; efforts will be made 
to make the second-generation vehicle smaller, 
lighter and more portable. It is the wish of the 
project team that the design benefit the widest 
possible range of disabled children. Discussions 
with several physical therapists from aroimd the 
country indicates that there are many children with 
abilities similar to Kramer (many spina bifida and 
cerd)ral palsy patients, for example) and other than 
the HPV described here, there appears to be 
nothing available for these children to ride. The 
goal of the second generation team is to aeate a 




O 




RESNA ’94 • June 17-22, 1994 



WHEELCHAIR LIFT CONTRAINDICATIONS 



F.L. Cardoso. M.S., J.G. Thacker. Ph.D. and A. Weinberg. M.S. 
University of Virginia Transportation Rehabilitation Engineering Center 
Charlottesville. VA USA 



Abstract 

The number of handicapped individuals 
depending on transportation assistive devices 
increases everyday, especially with the help of the 
Americans with Disabilities Act of 1990. The 
disabled population travels in a variety of 
wheelchairs. Many of these individuals seeking an 
independent lifestyle through transportation in 
personal vehicles also face the need for a wheelchair 
lift. They also often encounter the difficult tasks of 
selecting an appropriate lift that matches their 
specific wheelchair and vehicle. Problems arise 
because of the lack of easily accessible dimensional 
information that will help one make these selections. 

The objective of this project is to provide 
consumers, wheelchair lilt prcscribers. occupational 
therapists, and even manufacturers with a document 
to aid in matching lifts and wheelchairs and best 
fulfilling the needs of the disabled. One of the main 
obstacles to overcome is that of developing an 
objective but comprehensive method to educate and 
disseminate this information about devices available. 
The first problem lies in identifying the irntJor 
dimensional interferences between existing lilt 
designs and existing wheelchair designs. The 
second problem is understanding contraindications 
for wheelchair lilts in three major areas: safety, 
convenience, and reliabilit)'. By understanding these 
conditions, better decisions can be made prior to 
purchasing equipment. 

Background 

The idea for contraindications for 
wheelchair lifts was originally started by one of 
TREC's (Transportation Rehabilitation Engineering 
Center) staff members who also worked as an 
occupational therapist at the Woodrow Wilson 
Rehabilitation Center. It basically consisted of a 
preliminary list of observations about the interaction 
between wheelchair lifts and their users, focusing 
specifically on situations where the user's needs were 
not accomodated. It was noted that simple problems 
such as a steep safety flap (roll-stop) angle may be 
sufficient to prevent individuals on manual 
wheelchairs from using a lift. 

The number and variety of problc^2 



recognized grew with TREC's involvement in 
evaluating the SAE's Recommended Testing 
Procedures for wheelchair lifts in personally licensed 
vehicles. A representative cross-section of lift types 
currently in the market has been used in testing the 
proposed standards and as a byproduct, specific 
strengths and weaknesses of these different devices 
were found. 

This paper proposes the development and 
dissemination of a list of such lift contraindications 
and a method of objectively matching wheelchair 
and lift dimensional requirements through the use of 
a tabular matrix. 

Methods 

Lift Contraindications : As mentioned, the three 

main areas of concern when assessing wheelchair 
lifts are: safety, convenience, and reliability. In 
essence, these qualities found in lifts are all tied 
together. A reliable and efficient lift properly 
selected to accomodate the user and his/her lifestyle 
will most likely safely fulfill its purpose. Again, the 
contraindications should serve as precautionary 
information so that consumers may learn about the 
product their buying, what to expect, and how to 
qualify or disqualify lifts for their needs. At present, 
the contraindications are still presented in a "list" 
format and typical ones are shown in table 1. 

A further step in this part of the project will 
be disseminate the current list of contraindications to 
the public for comments, additions and 
recommendations. It is still unknown how to best 
provide this type of information for most efficient 
use and implementation. The separation of the 
contraindications into specific categories following a 
checklist used to match the lift with personal 
requirements may be a feasible possibility. 

Wheelchair and lift dimensio na) requirements: In 
many cases, the wheelchair or scooter is too long to 
fit on the lift in a safe manner. Either the wheels 
prevent the roll-stop barrier from locking into place 
or the or roll-stop flap interferes with the occupants 
feet. In other situations, the wheels of the 
wheelchair may get wedged onto the vehicle's side or 
bumper. Therefore safety and effectiveness are again 
major considerations. 

The main concerns of this part of the 



RESNA*94 • June 17-22, 1994 



79 



WHEELCHAIR LIFT CONTRAINDICATIONS 



• Platfor m lifts require - 8 ft, of space for exit from a lift. Rotary lifts need - 4 1/2 ft. 

• Manual Back-up Systems are often time consuming. Those used in hydraulic lifts including 

gravit>'-down operation of the platform are easier and faster to operate. 

• One-arm Folding Platform lifts tend to be less stable. 

• The Under-Vehicle Lift may present problems with dirt build-up and penetration. Its main 

advantage is that it does not take up interior space or block the van door. 

• Arm Lifts also allow for easy entry and do not block doorways. Applicability to dififerent 

wheelchairs is questionable since loads are directly applied to the wheelchair frame. 

• Mechanically or hydraulically interlocked roll-stops are more reliable and safer than spring 

loaded or gravity activated ones. 



Table 1: Major Lift Contraindications 



project were defining what wheelchair and lift 
dimensions to specify and how to present these 
guidelines in a simple format. Anyone interested in 
acquiring wheelchair and lift dimensions from 
manufacturers soon finds that each inaniifacliirer has 
its own way of defining dimensions and taking 
measurements. In an attempt to use a widely 
accepted method, TREC chose the ANSI/RESNA 
WC93-1991 Wheelchairs Maximum Overall 
Dimensions standard. Infonnation about a specific 
wheelchair model needs only two dimensions, the 
overall length and overall width, as shown in Figure 
1. The length of wheelchairs generally range 
between 1100 and 1300 mm. The width of 
wheelchairs are generally between 600 and 700 mm. 
It is assumed that the user's feet add approximately 
50 mm to the overall length. 

Using these collected dimensions and the 
inside dimensions for wheelchair lift platforms, also 
measured as shown in Figure 1, a cross-reference 



matrix for compatible lifts and wheelchairs will be 
created. Basically, the overall length of the 
wheelchair, L, plus a minimum clearance of 50 min 
should be less than or equal to the inside usable lift 
platfonn length. Likewise, the width of the 
wheelchair, B, should be smaller than the inside 
usable width of the platform with some mini mum 
clearance yet to be defined. Taking measurements in 
this manner will actually result in a slightly 
conservative fit since the foot rest and person's foot 
are assumed to be inside the platform usable area. 
This means that the footrest is not hanging out over 
the roll-stop flap, which occurs in some cases. 

The data on wheelchairs and lifts collected 
from the manufacturers is used to create a matrix 
like the sample on Figure 2 w'hich cross-references 
wheelchair and lift dimensional compatibility. Tliis 
quick reference table will serve as an immediate 
check for users and prescribers according to the 
measurement guidelines. 




Figure 1: Appropriate wheelchair and lift measurements for these guidelines. 

- * 94 

RESNA ’94 • June 17-22, 1994 



80 



WHEELCHAIR LIFT CONTRAINDICATIONS 



LIFTS 

wii£Eix::mairs 

and SCOOTERS 


Rlcon 


Crow River 


Braun 




stooo 

29W X 42L 


SlOOl 
29VV X 44L 


7688LA 
29W X 42L 


Vangatcr 
30W X 42L 


u 

'T 

M 

1 t 




-L 


Invocure 


lOOOE 

25.5W X 36L 


✓ 


✓ 


✓ 


✓ 


✓ 




\ 


1000 

25.5W x32L 


✓ 


✓ 


✓ 


✓ 


✓ 






E.J. 


Universal 
22.SW x36L 


✓ 


✓ 


✓ 


✓ 


✓ 






VUta 

22.75W X 31L 


✓ 


✓ 


✓ 


✓ 


✓ 




J 


Fortre** (ncooler) 


2000FS 
24W X 43L 


X 


✓ 


X 


X 


X 


































LJ 



Figure 2; Sample schematic of the matrix of wheelchair / lifts reference. (The data shown is 
hypothetical and does not represent actual values) 



Discussion 

Although some iufornuHion can be found in 
related literature, it is believed that occupational 
therapists and prescribers must rely mostly on their 
personal and professional experiences and their 
locally accepted guidelines when matching lifts with 
the requirements of their handicapped clients. An 
article recently published by Perr and Barnicle (3) 
accurately mentioned how complicated selecting the 
appropriate lift can be due to the variety of lifts and 
wheelchairs available, diversity of special needs oi 
the disabled user and lack of information. 

This project aims at making proper 
wheelchair and wheelchair lift selection easier for 
consumers and prescribers. By providing an 
objective document with information on 
contraindications for wheelchair lifts and what 
characteristics to look out for when matching 
personal needs with available devices, TREC hopes 
to educate people so that better decisions can be 
made. In order for this document to achieve its 
objectives, it will have to be reviewed and updated 
periodically, especially by active users and 
prescribers. Compatibility between the vehicle, lift, 
wheelchair and user's requirements is essential in 
making this technology work towards a safe and 
nonnal lifestyle. 



Acknowledgements 

Funding for research was provided by the 

National Institute on Disability and Rehabilitation 

Research grant Number H133E00006. 

References 

1. Lifts for Vans : REquest, National Rehabilitation 
Hospital, Washington D.C., December 1992. 

2. Carpenter, Dale. "The Intex FX700 Wheelchair 
Lift - Design Considerations and Results"; SAE 
Paper No. 851663. Warrendale PA, Society of 
Automotive Engineers, 1985, p 7-17. 

3. Perr, Anita, Barnicle, Kitch. "Van Lifts: The 
Ups and Downs and Ins and Outs"; Team 
Rehab Report, June 1993, p. 49-53. 

4. American National Standard ANSI/RESNA 
WC93-1991, Wheelchairs Maximum Overall 
Dimensions, RESNA, Washington D.C., July 
30, 1991. 



BEST COPY AVAILABLE 



RESNA ’94 • June 17-22, 1994 



81 



POSITIONING AND SECUREMENT OF RIDERS AND 
lilEIR MOBILITY AIDS IN TRANSIT VEHICLES 



Patricia Karg, Kirsten Yalfe 
BCRI, Plymouth Meeting, PA 



ABSTRACT 

Individuals are transported daily in transit vehicles 
while seated on their mobility aids. Although 
federal standards exist to regulate this practice, 
there is controversy over whether the regulations 
are effective or practical. This paper describes the 
results of a study that performed an objective 
review of information available in several areas 
relevant to positioning and securing riders and 
their mobility aids in transit vehicles. The 
objective of the study was to present the current 
status of knowledge and to identify gaps in 
knowledge and related problems, As a result, 
recommendations are provided on fbture research 
needs and options for improvement of policy. 

INTRODUCTION 

Approximately two-thirds of an estimated 1.5 
million mobility aid users require special 
transportation services (La Plante 1992). 

According to a recent profile of public transit 
passengers nationwide, an average of 1.2% of all 
transit riders have disabilities; the percentage may 
be 10%-15% in some communities (APTA 1992). 

Accessible transportation allows people with 
disabilities to commute to work and school and to 
take part in normal daily activities. Accessible 
transportation includes various types of transit 
systems, such as fixed-route, demand-responsive, 
and paratransit services. 

The 1990 passage of the Americans with 
Disabilities Act (ADA) is intended to help ensure 
adequate access, but the act does not specify the 
details of implementation. A number of 
controversial issues remain unresolved, including 
the best way or ways to provide safe, effective, 
and comfortable positiomne and securement of 
riders and their mobility aids in transit vehicles. 
Much of the debate focuses on whether to position 
mobility aids in a forward-facing, rearward-facing, 
or angled securement system. 

Federal regulations for transit vehicles currently 
mandate that the orientation of at least one of the 
required securement systems be forward facing; 
the remainder mw be rearwwd facing with a 
padded barrier. Other orientations of securement 
systems are currently prohibited. 

Some transit providers and consumers have 
challenged this ban on side-facing and other 
angled orientations, arming that mese orientations 
may be as safe with all factors considered (e.g., 
the use environment of the vehicle) and may 
provide other advantages, as well. Some 
providers believe that forward or rearward f^ng; 



reduces vehicle capacity, resulting in a loss in 
quality and an increased cost of service. 
Consumers have also voiced concerns about ease 
of emergency evacuation, passenger comfort, and 
the users' ability to secure themselves unaided 
when in a forward- or rearward-facing position. 

A number of potential information sources may 
facilitate an understanding of this issue, including 
legislation, regulations, and standards: accident 
statistics; and oiomechanics and physical (crash) 
testing research. In addition, transit providers, 
consumers (mobility aid passengers), device 
manufacturers, and government agencies have 
various experiences and opinions. 

The development of satisfactory solutions is 
hindered, however, because the knowledge base is 
distributed among many individuals and 
organizations. Overall system perspectives are not 
always readily apparent due to differences in 
objectives and compaitmentalization of 
iniormation. Information critical to this issue, 
though readily available, may be overlooked 
during studies because the group sponsoring or 
performing the study does not recognize this need. 

This paper describes a study that performed an 
objective review of information available in 
several areas relevant to the positioning and 
securement of mobility aids in transit vehicles. 

The objective of the study was to present the 
current status of knowledge, identity gaps in 
knowledge and related problems, and recommend 
future research needs and policy direction. The 
scope of the project was limited to consideration 
of issues in vans, and small and large transit 
buses. Although school buses were not evaluated, 
information on these vehicles was collected and 
applied to vehicles within the project’s scope. 

METHODS 

Irtformathn Sources 

A panel of experts were drawn from 
representatives of government agencies, transit 
systems, consumer organizations, manufacturers, 
test facilities, associations, and others. The Project 
Review Conunlttee (PRC) brought together 
expertise on the technical aspects of equipment 
and biomechanics, operational aspects of transit 
systems and consumer concerns; and knowiedge 
of regulations, safety, and risk management issues. 

An extensive literature search was performed. In 
addition, we collected information on 
commercially available mobility aid securement 
systems to determine the state of the art of this 
technology. We obtained information through 



82 



RESNA ’94 • June 17-22, 1994 



Positioning & Securement: Transit Vehicles 



personal contacts with PRC members and transit 
providers at conferences and meetings. 

A survey of transit systems was developed to 

E rovide current, useful information that could not 
e ascertained from the literature. We designed 
the survey to provide accident data, transit fleet 
characteristics, and current practices for securing 
mobility aids and their users. 

We developed a database to track and organize all 
documents gathered during the study. Each 
document in the database is classified according to 
the type of information it contains (c.g., 
documentation from crash testing, a review of 
literature, standards, regulations, conference 
proceedings); the database also includes abstracts 
tor several documents. In addition, each 
document record contains 29 standardized topic 
fields. For each topic addressed in the document, 
the corresponding neld is indicated. Users 
search the database for all documents that address 
a particular topic; the database is intended to aid 
in future research efforts, as well. 

Criteria and Method of Analysis 

Based on the literature, our discussions with 
individuals and organizations, a survey of the 
experts in our steering conunittee and of tr^sit 
providers, the following criteria are generally 
accepted as being important when designing and 
assessing securement systems. In practice, there 
will need to be tradeoffs between criteria, so we 
have listed the criteria in order of priority ^d 
indicated the general category into which they tall. 



1. Minimize injury to the occupant during 
impacts. Minimize injury to the occupant 
during normal driving and emergency 
maneuvers (e.g, swerving, hard braking). 

2. Allow for egress in an emergency 
situation. 

3. Simplify use of securement and passenger 
restraint devices. (If the devices are 
difficult to use, passengers and/or operators 
may not use them or use them improperly.) 

Ersonomics/Human Factors: , 

4 Simplify use of securement devices by the 
operator and/or passenger, to allow for 
independent and timely securement. 

5 Provide for the mobility aid users physical 
and mental comfort. For example, some 
orientations may cause nausea for some 
users or place diem so that they are facing 
direcdy at other passengers. 

Vehicle Characteristics: 

6. Optimize vehicle capacity for conventional 
seatii^ and mobility aid passengers. This 
significantly impacts the need for 
additional vehicles, scheduling, and cost. 

7. Minimize costs (e.g., securement systems, 
installation). 

Data were collected, organized, and analyzed in 



the areas of accident statistics, U.S. regulations, 
internationd standards and guidelines, physical 
testing research, biomechanics research, vehicle 
characteristics, ergonomics, and human factors. 
These areas were chosen based on a review of the 
literature and the concerns expressed by transit 
providers, consumers, researchers, and regulators. 

SUMMARY OF FINDINGS 

The main results of our analysis were: 

• Existing accident data do not reveal the 
effect of mobility aid orientation on transit 
passenger safety. For example, the data do 
not conclusively indicate the predominant 
direction of impact for transit vehicles. 

The methods of data acquisition are neither 
uniform nor comprehensive and, therefore, 
cannot currently be used for the purpose of 
assessing the relative safety of various 
seating orientations. 

Limited data on injury-causing events for 
transit passengers riding on their mobility 
aids showed injuries to be primarily due to 
improper securement during normal and 
emergency driving conditions. 

• U.S. regulations and most iritemational 
standards require that mobility aid 
securement/occupant restraint systems be 
oriented forward or rearward with respect 
to the direction of travel and meet 
specified performance requirements. 
Requirements for systems’ performance 
specify criteria under frontal loading 
conditions. Testing under lateral loading 
has not been developed. 

• The physical testing of mobility aid 
securement and occupant restraint systems 
that was the basis for ruling out side-facing 
orientations was not intended to evaluate 
the best way to position passengers and 
their mobility aids. Qualitative analysis of 
test data showed that the systems tested 
provide inadequate protection when the 
mobility aid is in a side-facing orientation 
with respect to the direction of impact. 

Our combined analysis of the injury and 
excursion data from several physical 
testing programs did not reveal benefits of 
one orientation over another. The analysis 
contraindicates carrying out a testing 
program, at this time modeled after these 
research programs, to answer the question 
of how best to position a mobility aid. 

• Information collected on passenger 
biomechanics and human tolerance limits 
indicates that the occupant is at a greater 
risk of injury in the side-facing orientation. 
Fronts impacts result in the most severe 
crash forces and the human body has a 



97 



r COPY AVAILABLE 



RESNA ’94 



June 17-22, 1994 



83 



Positioning & Securement: Transit Vehicles 



reduced tolerance to lateral loads. 

• Research and development efforts in 
occupant protection nave been geared 
toward protecting passengers in the 
forward-facing orientation. As a result, 
product development of mobility aid 
securement ana occupant restraint systems 
has focused on achieving safe^ in this 
position under extreme conditions of 
frontal impacts. Lateral stability and injury 
prevention has not been develo^d. 

• Capacity reduction in transit vehicles 
appears to affect smaller vehicles and to be 
a result of compliance with a variety of 
ADA requirements. It does not appear to 
be due to orientation requirements alone. 
Research into optimal vehicle layouts and 
use of innovative seating technology may 
reduce capacity problems. 

• No collected data indicated a choice of 
orientation based on ergonomic and human 
factors. Once an orientation has been 
chosen, innovation in vehicle layout and 
securement system design should address 
and resolve problems in these areas. 

• An indirect conclusion is that there is a 
need for development of a procedure to 
determine if ceitain mobility aids should be 
occupied during vehicle transport. 

CONCLUSIONS AND RECOMMENDATIONS 

Research Issues: 

1. Do not pursue a research program 
consisting of sled testing of angled 
mobility aid orientations (including 90°) at 
this time. 

2. Support research into improved securement 
systems for mobility aids. 

3. Support further research and development 
of testing programs that qualify mobility 
aids for use on transit vehicles. 

4. Pursue the development of test methods 
and criteria for testing side-facing or other 
angled securement and occupant restraint 
systems for equivalency and for testing the 
stability and protection of forward- or 
rearward-facing systems under lateral loads 
(e.g., swerving, side impacts). 

5. Pursue research into optimal vehicle 
layouts and use of alternative seating 
technology in vans and small buses to meet 
the needs of transit providers. 

6. Support efforts to gather better transit 
accident/injury data prospectively for 
transit incidents through development of a 
database for problem reporting. 



Policy Issues: 

1. For the present, support the policy of 
requiring forward- or rearward-facing 
mobility aid securement systems based on 
the results of the collective analysis of the 
data. Sufficient data do not exist to show 
that an angled orientation (including side 
facing) is equivalent in safety to the 
forward- or rearward-facing orientations or 
that it will provide significant advantages. 

2. Examine the policy requiring transit 
vehicles to transport riders seated in all 
“conunon” mobility aids. Until further 
research and device improvements are 
performed, not all mobility aids can be 
transported safely while occupied. 

In summary, the study recommended that, for the 
time being, the regulations should continue to 
require forward- or rearward-facing securement 
systems. At the same time, it recommends 
development of performance criteria and test 
methods that will ensure the safety of 
securement/occupant restraint systems under side- 
loading conditions, allowing for innovation and 
improved safety. 

The study further recommends that research must 
be performed in the areas of mobility aids, vehicle 
layouts and seating systems, and 
securement/occupant restraint systems and the 
procedures used to test them to aid transit systems 
in implementing the ADA without compromising 
safety or the quality of service. In addition, 
creation of a problem reporting system would be 
useful for assessing the needs for improvement in 
technology and transportation procedures. 

REFERENCES 

American Public Transit Association (APTA). 
Americans in Transit~A Profile of Public Transit 
Passengers, Technical Services Department and 
Research and Statistics Division, Washington, DC, 
1992 Dec. 

La Plante MP. People with disabilities in basic life 
activities in the U.S. Disability Statistics Abstract, 
No, 3. Disability Statistics Program, Institute for 
Health and Aging, School of Nursing, University 
of California San Francisco, 1992 Apr. 

ACKNOWLEDGEMENTS 

Funding for this project was derived from the 
Federal Transit Act, as amended, through a 
Cooperative Agreement with the U.S. Department 
of Transportation, Federal Transit Administration 
and Project ACTION of the National Easter Seal 
Society. 

ECRI 

5200 Butler Pike 

Plymouth Meeting, PA 19462 

(610)825-6000 



ERIC 



84 



^ 98 

RESNA’94 • June 17-22, 1994 



t V I'i t W ti* 



SIG-03 

Augmentative & Alternative 
Communication 



A COMMUNICATION SYSTEM ESTABLISHED FOR A C-5 QUADRAPLEGIC AND 

HIS HEARING IMPAIRED PARENTS 

Deborah L. Rosenblatt, OTR/L 
Shepherd Spinal Center, Atlanta, Geoi^ia 



Abstract 

A young man loses the ability to co mmuni cate 
basic needs and emergency situations to his 
hearing impaired parents secondary to his high 
level spinal cord injury and closed head injury. 
The patient’s rehab team considered this 
situation unsafe for the patient to return to his 
home environment since he had several medical 
complications. This paper examines the special 
needs of this patient and his &mily and 
describes how the team was able to establish a 
safe home environment for this young man once 
he was discharged from the hospital. 

Background 

Medical Historv/Psvchosocial Status 
John, a 17 year old male, was rendered 
quadriplegic and sustained a closed head injury 
following a motor vehicle accident on 1/1 1/93. 
I nitiall y , the patient was able to open his eyes 
and follow directions and move the u{^r 
extremities. On 1/21/93, the patient underwent 
an emergency tracheostonty secondary to 
desaturation, braitycardia and a seizure. The 
patient was noted to have a neurological decline 
secondary to unresponsiveness and less 
movement in the upper extremities. The patient 
was transferred to Shepherd Spinal Center on 
2/17/93 where he was diagnosed with C-5 
complete quadriplegia and TBI status post- 
anoxic event. One other complicating factor in 
John’s admission was his Grade IV sacral 
decubitus which required surgical intervention. 
John's rehabilitation course was delved 
secondary to his confinement to the bed for eight 
weeks following surgery. 

John’s parents are both completely deaf and his 
father is visually impaired as well. Premorbidly, 
John communicated with his parents through 
sign language. John also has'an older sister who 
has normal hearing. John's sister did not live at 
home at the time of the accident but moved back 
home after the accident. The sister stated that 
she did not want to be responsible for all of 
John's care once he came home since she 
worked and had a life of her own. The famil y 



has limited financial resources and lives in a 
first floor, three bedroom apartment in a rural 
town. John's grandmother, aunt and cousins 
live in the same town, although are not within 
walking distance to John’s apartment. 

John attends school at the local high school 
approximately 3 miles from his home. In 
S^ember of 1993, he will be in the eleventh 
grade. John’s personal goal was to return to 
school. John was considered a good student 
prior to his accident and was very interested in 
pursuing art. John was a talented artist and 
many of his drawings were displ^ed in his 
hospital room. Other interests include r eadin g 
and watching movies and sports events on 
television with his &ther. 

Problems 

Acutely, John was not able to speak secondary to 
his trach tube. He also exhibited severe 
dysarthria initially as a result of his head injury 
and was unable to even mouth words. John 
became very frustrated with his difftculty 
communicating with others, especially his 
parents. John was paralyzed below his 
shoulders and had very limited range of motion 
in his shoulders and neck secondary to severe 
tone. It was therefore impossible for John to use 
conventional sign language to co mmunica te 
with his parents. 

Although John regained the use of his voice and 
speech sufficient for hearing persons, his 
inability to communicate with his parents raised 
the question of John’s ability to return home and 
live safely with his parents. John was prone to 
autonomic dysreflexia secondary to his high 
level of spinal cord injury. This condition is life 
threatening and needs immediate attenri nn 
John was also not able to control selected 
environmental functions secondary to his 
physical limita ti ons 

John was dependent for all functions that take 
place in the school setting. 




• June 17-22, 1994 



RESNA ’94 



86 



Obiective 

John's rehabilitation team met weekly to discuss 
John's goals, problems and progress. The rehab 
team consisted of the pediatrician, nurse, 
occupational therapist, physical therapist, 
speech pathologist, social worker, school 
teacher, psychologist and rehab technologist. 
During these meetings, John's discharge plans 
were discussed and decided. Many times the 
issue of whether or not John's parents would be 
able to safely care for John at home was 
discussed. John’s parents were not able to hear 
John if he called for help and they were not able 
to understand what he needed at any given time. 
It the mission of the team to establish 

systems in the home that would enable John and 
his parents to effectively communicate and to 
decrease the potential for problems to a 
minimum. TTie team also was interested in 
getting John up with systems that would enable 
him to return to his school so that he could 
finish high school and pursue college. 

Method/Approach 

John was referred to an Assistive Technology 
Team by his occupational therapist in order to 
evaluate and e>q)erience different systems that 
would meet the needs of both John and his 
parents. The Assistive Technology Team, 
consisting of a speech pathologist, rehabilitation 
technologist, and assistive technology 
occupational therapist, evaluated John for a 
system that he could easily utilize when he 
needed to get his parent's attention or in the case 
of an emergency. Components of the system are 
listed below. 

Se ating and Mobilitv 

John was positioned in a power tilt in space 
wheelchair to prevent shearing on his sacral 
area and to prevent increased tone when doing 
mandatory weight shifts every 30 minutes. John 
was eventually able to develop enough strength 
in his right upper extremity to be aWe to drive 
his wheelchair with a hand control. Prior to 
using the hand control, John had become 
proficient in using sip and puff to control the 
chair. A Plexiglas laptr^ was also prescribed 
for John to support ecpiipment and schoolwork. 

Emergency call system 



RESNA ’94 



John used sip and puff switch access since he 
had limited use ofbilateralrqjper extremities. It 
W3S recommended that John have an emergency 
call system that flashed lights in the house and 
activated a bed vibrator so that his parents 
would respond to the call. An X-10 based 
system was recommended. John was issued the 
XIO transmitter/receiver, a sip and puff 
pneumatic switch with a gooseneck mounting 
system, 5 aR)liance modules, 5 flasher buttons 
to make the lights flash, and a bed vibrator. 

John was also issued a q)eaker phone adaj:rted 
with a sip/puff switch to allow limited 

telephone operation. The speaker 
phone could be used as a backup system in case 
his parents could not understand what he 
needed. The speaker phone had 911 
preprogrammed so that it would be easy for John 
to contact emergency help. 

John's rehab team made a home visit to John's 
house in order to set up the equipment and to 
train the flunily on its use. Eturing this visit, 
calls were made to the local police and fire 
department and to the local hospital to let them 
know that John would be returning home. 

rnmmiinication Needs 
John was in need of a commumcation system 
that would allow him to communicate with his 
parents in a consistent manner. After many trial 
and error sessions in occupational therapy, John 
demonstrated his proficiency with the Prentke- 
Romich Head Master as a computer access tool. 
This device operates the cursor with head 
movements and sip and jniff. The team then set 
out to find a computer program that could be 
iirili7fld with a pOTtable computer that would 
allow John to spell out what he wanted to say 
and also display pictures of what he was 
expressing. The pictogr^hs were necessary as 
John's father had difficulty seeing the computer 
letters and both parents have low educational 
levels. John's speech pathologist recommended 
the computer program "Speaking Dynamically" 
for Macintosh computers after review of several 
programs. 

An /^le Macintosh PowerBook 180 laptop 
computer was also recommended for John. It 
was imp ortant that this specific computer be 
recommended as the technology team evaluated 
the readability of the screens for both pointer 
visibility when used with the Head Master and 

Ibi 

June 17-22, 1994 



87 



also visibibty with John's mother and father. 
This computer was the only one that provided 
active matrix black and white displays that 
approach the quality needed for contrast, size 
and angle viewing. John and his parents 
underwent intensive &mily tr ainin g to learn this 
Communication system. It was necessary to hire 
a sign language interpreter to help teach the 
parents the system. The &mily participated in 
training daily for one and a h^ hours per d^, 5 
days a week for 2 and a half weeks. T raining 
continued until both parents could demonstrate 
their ability to correctly and consistently use the 
communication system Eventually, both John 
and his parents were able to prove then- 
understanding of the system when they 
completed a therapeutic pass inside the hospital 
for 48 hours with no assistance from the staff in 
John’s care. 

A Daessy Folding Mount was also reconunended 
so that the laptop computer could be mounted to 
John's wheelchair. A power inverter was 
recommended to allow the portable conqniter to 
be tied to the wheelchair batteries fijr extended 
periods of operation. 

School 

In order for John to return to school, he needed 
to be able to use a computer to complete his 
assignments. His laptop computer can easily 
meet his needs with the following software 
additions. 

■ScreenDoors 1 . 1 for the Macintosh using 
System?.; This software program emulates a 
keyboard on the screen and has word predictioiL 
This allows John to type his p^rs with the 
HeadMaster. 

-Claris Works 2.0: This software program 
includes a word processor, spreadsheet, 
database, telecommunication and drawing 
functions. 

■Professor Mac; A tutorial software program for 
the Apple Macintosh. 

A HP Deskwnter for Mac was the recommended 
printer and a Practical 24/96 Fax Modem was 
also recommended so that John could easily 
print out his wort or could fex his work to the 
school if he became confined to his home for 
any reason. 

Funding 



Funding for the recommended equipment 
became a real issue when John was ready to be 
discharged. John’s insurance coverage was 
limited and standard and would not cover his 
very specialized equipment needs. The rehab 
team applied for and obtained a private gr ant 
that funded the equipment that would enable 
John to be safe in his home. John’s high school 
held a fund-raiser that raised the r emaining 
fiinds necessary to obtain John’s school 
equipment. 

Results 

John was discharged home with his parents on 
August 4, 1993 with plans to return to school at 
the end of August. Since John's discharge, 

John's rehab team has been communicating with 
his school to inform them of his needs in the 
school setting. John was also referred to 
occiqiational therapy in the school so that he 
could receive further computer training once he 
received his software. 

Discussion 

The integration of a variety of environmental 
control units, a computer and corrqxiter program 
and a method of access, has allowed a young 
man to return home with his fomily and return 
to school. Marty hours of team prcfolem solving, 
eiqierimentation and training h^ to occur in 
order to allow this to happen. The availability 
and corrqratibility of certain types of technology 
allowed the team to come up with a solution that 
at one time seemed impossible. 

Manufacturers 

a. Radio Shack 

500 One Tanrty Center 
Fort Worth, TX 76102 

b. Prentke-Romich Company 
1022 Heyl Road 
Wooster, OH 44691 

c. Shepherd Spinal Center 

Department of Rehabilitation Technology 
2020 Peachtree Rd, NW 
Atlanta, GA 30309 

Address 

Deborah L. Rosenblatt, OTR 
Shepherd Spinal Center ^ 

2020 Peachtree Road NW. 

Atlanta, Georgia 30309 

assir copy avaiubli 



102 

RESNA ’94 June 17-22, 1994 



88 



DEVELOPMENT OF AN ASSISTIVE TECHNOLOGY SUPPORT NETWORK 



Albert M. Cook, Susan M. Hussey and Colette L. Coleman 

Assistive Device Center 
California State University, Sacramento 



er|c 



Abstract 

The goal of this project 
was to increase opportun- 
ities for interaction, 
control and communication 
by adults with develop- 
mental disabilities in 
both their day program and 
group home settings. An 
integrated program of 
community-based service 
delivery, staff training 
and consultation was 
developed to meet these 
needs . 



Rackaround 

This project developed a 
community-based assistive 
technology intervention model 
for the improvement of 
interaction skills in 
severely developmental ly 
disabled adults. 

The project was divided into 
two phases. During the start 
up phase, two group homes 
were used as field sites for 
development of the community- 
based assistive technology 
intervention approach and 
model. During the transition 
phase of the project two 
additional group homes were 
selected for replication of 
the program. 



Object ive 

The overall goal of this 
project was to develop a model 
program for development of 
local AT support networks 
which facilitate interaction 
and communication for adults 
who live in residential care 



RESNA’94 • 



facilities and attend day 
programs. Specific objectives 
were to: (1) identify and 
describe a framework for cost 
effective community-based AT 
intervention, (2) develop 
program guidelines that can 
be used for replication, (3) 
develop a training program 
and materials for training 
direct care staff, (4) develop 
program materials for 
consumer training. 

Method/Approach 
During the start-up phase we 
developed a collaborative 
consultation/training model 
for intervention. This model 
included a large training 
component and materials were 
developed. These resource 
materials included suggested 
equipment lists, recommended 
activities for participant 
training, and data collection 
procedures and forms. Train- 
ing sessions were videotaped 
for use in developing 
materials and for use with 
new staff. Further training 
of direct care staff occurred 
during collaborative consumer 
training sessions. A set of 
forms for assessment, 
intervention planning and 
monitoring progress were also 
developed. Equipment utilized 
in the training was supplied 
to each group home by the 
project. Within each group 
home and day program a 
resource person was 
identified. This was essen- 
tial for establishing lines 
of communication and the 
level of support required by 
direct care staff. 

Juril^'^22, 1994 89 



AT Support Network 



Both formative and summative 
evaluation were employed to 
evaluate the impact of the 
project on the assistive 
technology skills of the 
direct-care staff and the 
program participants. A 
pre-test addressed initial 
assistive technology know- 
ledge and attitudes, as well 
as perceptions of consumers 
interactive skills by the 
staff. Following staff and 
consumer training, post-tests 
were utilized to assess 
learning. Direct-care staff 
also provided informal 
evaluation of all project, 
activities . 

Low-technology approaches 
were also included through 
consultation on the design 
and development of 
communication boards which can 
be utilized by consumers. Our 
materials included an 
environmental communication 
survey which identified 
communication opportunities 
for participants and resulted 
in suggestions for specific 
vocabulary to be included on 
communication boards. 

As the program developed, and 
consumers and staff developed 
greater AT skills, it was 
necessary to set new goals 
and make adjustments in the 
training programs. 

During the transition phase, 
we replicated the intervention 
model at two additional group 
homes. This phase allowed us 
to evaluate our model and 
materials in a mode very 
similar to that which will 
exist during the post-project 
continuation phase. During 
this phase, consultation was 
provided in at least one., 
session per month per group 



home. Direct care staff 
provide on-going training of 
consumers between these 
sessions . 



Results 

Training materials were 
developed for three audiences: 
(1) professional consultants 
to the group home or day 
program (speech pathologist, 
occupational therapist), (2) 
direct care staff and (3) 
consumers. For the first two 
of these competencies were 
developed. For each consumer 
intervention goals were 
established. 

The pre-test results indicated 
a general lack of familiarity 
and comfort by staff with 
computers and other 
technologies. Following 
training, the staff became 
more comfortable with the 
computer, and they were able 
to independently use it and 
the other technology with the 
consumers in their programs. 

Discussion 

Personal computer-based and 
low-technology equipment for 
training and evaluation was 
used. Since it was possible 
to simulate participant 
interaction with an individual 
communication device using the 
computer, this approach 
allowed determination of 
whether a specific device 
would be successful for a 
given individual. Since the 
training and evaluation can 
occur over a prolonged time 
period, this approach gave the 
consumer a "fair chance" at 
achieving success without the 
purchase of an individual 
device. Since funding agency 
funds won't be expended for 
an individual device until 
successful use is demonstrated 



104 

RESNA ’94 • June 17-22, 1994 



90 



AT Support Network 

based on a computer 
simulation, costly "mistakes" 
in equipment acquisition can 
be avoided, and the likelihood 
of success with the individual 
device which is eventually 
chosen will be greater. 



The consumers participating in 
this project improved both 
their control of the 
environment and their ability 
to communicate. Both of these 
lead to greater participation 
in daily activities and an 
intensified role in self- 
determination. 

The perceptions of the direct 
care staff towards technolgy 
and its benefit for the 
consumers also changed 
significantly. 

Acknowledgements 

This project was funded by the 

Targeted Program Development 

Fund of the California State 

Department of Developmental 

Disabilities. 

Appendix 

Equipment utilized- Computer 
system: MacintoshLC w/ApplellE 
card, Ke:nx , Echo LC speech 
synthesizer, color monitor. 
Control interfaces: enlarged 
keyboards^ switches : tread 
and lever , wobble*, big red 
and jelly bean% universal 
switch mounting^; Ablenet 
Appliance Control Unit, 
battery adapters , single 
switch Ultra-Four transmitter/ 
receiver pairs . 



Albert M. Cook, Ph.D. 
Assistive Device Center 
California State University 
6000 Jay Street 



^Don Johnston Developmental 
Systems 

^Intellitools , Inc. 

^Zygo Industries 
^Prentke Romich Company 
^Ablenet Inc. 

®Tash, Inc. 



105 

RESNA’94 • June 17-22, 1994 91 



A HEAD GESTURE RECOGNITION SYSTEM FOR COMPUTER ACCESS 



Costi Perricos and Robin D Jackson 

Department of Engineering 
University of Cambridge, UK 



ABSTRACT 

This paper describes woric done in the recognition of 
head gestures at Cambridge University. The aim of 
this work is to enable quadriplegics and individuals 
with various muscular disorders to interact with 
computer applications and to communicate with their 
environment. The approach to the problem uses 
template - based recognition techniques to achieve 
real - time recognition of a gesture vocabulary of 
variable size. The calibration of the recogniser is 
achieved by using a software package, which 
optimises the procedures for an individual. These 
procedures can then easily be incorporated into any 
application involving a human - computer interface. 
Preliminary experimental results obtained from trials 
with subjects who have athetoid cerebral palsy are 
presented. 

BACKGROUND 

Certain individuals suffer from severe handicaps such 
as high - level quadriplegia, cerebral palsy and other 
neuromuscular disorders, which restricts their access 
to computer - based applications or communication 
aids and limits their capability to control their 
environment. This limited communication can have 
traumatic psychological effects on both adults and 
children, and the lack of manipulative ability may 
prevent children from following the natural learning 
process. Typical computer access methods include 
keyboards, joysticks, mice and touch pads, all of 
which are useless to users without some manual 
dexterity. Speech recognition systems offer a partial 
solution to the problem, but some severely 
handicapped users are dysarthric and can therefore 
satisfactorily control only their head muscles . 

Substantial research work has been done in the 
recognition of head gestures, using techniques such as 
hidden Markov models, neural networks, and finite 
state machines[JL][2][2]- These procedures have 
generally been tailored to particular applications, and 
have not been flexible or easily accommodating to any 
changes in the user's ability to perform head gestures. 

This paper describes the design and development of a 
gesture recognition system that can run on a 386 or 
486 based personal computer. The training of the 



recogniser is simple and fast, and once trained can be 
incorporated in any computer application being 
developed. The set-up used in the user trials consisted 
of a 386DX based computer with a 387 maths 
coprocessor. A Polhemus 3Space Isotrack transducer 
mounted on a baseball cap was use to measure the 
head position. The decision to use the Polhemus was 
one of availability, but similar performance would be 
expected from a cheaper head position sensor. 

PROBLEM STATEMENT 

In order to design an effective gesture recognition 
system, certain design requirements had to be 
fulfilled. These can be summarised as follows: 

• The system must be simple to use. 

• It must be inexpensive. 

• It must have the ability to adapt to an 
improvement or deterioration of the user's ability 
to perform head gestures. 

• Its gesture recognition accuracy must be high. 

• The system should be trainable on small amounts 
of data. 

• It should be able to cope with a range of gesture 
vocabularies. 

• It should be easy to incorporate in the 
development of computer applications. 

It is important to note that due to the large variation 
in disability type and severity, it is practically 
impossible to design a user independent recogniser. It 
is possible, however, to minimise the variables needed 
to define the characteristics of the recogniser, and to 
make its calibration simple and user friendly. 
Furthermore, although the recogniser should cope 
with a range of gesture vocabularies, it was thought 
useful to define a base vocabulary of gestures. This 
has the advantage of providing software developers 
with a minimum vocabulary of gestures to incorporate 
into their applications, as well as establishing a 
minimum level of ability required to use such a 
system. The gestures used in the base vocabulary 
should be intuitive and easy to perform. A set of six 
simple gestures was hence chosen, namely Up, Down, 
Left, Right, Yes and No. In addition to these simple 
directional' gestures, the user can choose oth» 
gestures to add to his vocabulary. These gestures are 
described as complex, and the current limit is set to 



RESNA ’94 • June 17-22, 1994 



92 



HEAD GESTURE RECOGNITION 



nine additional complex gestures, making a total 
vocabulary of 15. 

DESIGN AND DEVELOPMENT 



simply replacing the gesture templates. In this 
particular case, a process known as Dynamic Time 
Warping (DTW) was used, and satisfactory results 
were obtained from Only one template per gesture. 



The gesture recognition process can be defined as 
follows; 



Incomin 

Gesture 



Movement 

Digitization 



Gesture 

Segmentation 



Classification . 

& -f Result 
Recogniuon 



The head movement produced is digitised using a 
Polhemus 3 Space Isotrak transducer sampling at 
60Hz. It provides the computer with a six - 
dimensional data vector containing three translational 
parameters and three rotational parameters of a head 
mounted sensor with respect to a stationary source. 
The gesture data is communicated to the computer in 
serial form. 



Once the head movement information has been 
digitised, the gestures need to be segmented, in order 
to define their start and end points in time. 
Segmentation of the gestures is performed by using a 
tremor filter. The tremor filter used in the recogniser 
consists of a first in first out (FIFO) buffer of fixed 
length which is updated after every sample. The 
tremor in each axis is measured by taking the 
variance of the head position over the time span 
defined by the buffer length. There are two tremor 
thresholds which are implemented in the 
segmentation process, namely the still threshold and 
the movement threshold. The still threshold 
determines when a person's head has gone from a 
still state to a moving state, while the movement 
threshold determines when the head has returned to 
the still state after having completed a gesture. Both 
thresholds are user dependent and are determined 
probabilistically by the calibration software, so that 
in each head state, the probability of the 
segmentation algorithm being correct is 95%. 



DTW has successfully been used in speech 
recognition, and the recognition of hand gestures (^. 
It is an optimisation technique whereby the vector 
distance between two finite, time varying signals is 
calculated, after the difference due to varying signal 
lengths has been minimised. Waibel and Lee provide 
a comprehensive reference on DTW [5]. 

A process known as principal coefficient analysis 
(PCA) is used to classify the gestures, before they are 
recognised by the DTW algorithm. This has the 
advantage of reducing the number of templates the 
DTW algorithm needs to search through, thereby 
increasing the recognition speed. PCA performs the 
following operation on the gesture data representing 
the three rotational axes of the head; 

= ( 1 ) 

Where: is the principal coefficient of axis ax 

is data point ax in the ith vector of the 
gesture sequence 

X is the mean value of x^j^ *n the vector 
sequence 

Since directional head gestures are dominated by 
movement in one axis of rotation, it is possible to 
distinguish them from complex gestures by looking 
at the principal coefficients obtained. The principal 
coefficient analyser performs this task by examining 
the Principal Coefficient Factor (PCF). which is 
defined as; 



PC 

PCF= (2) 

PQ+PQ 



When considering which recognition techniques to 
use, we examined stochastic and connectionist 
recognition techniques such as neural networks and 
hidden Maiicov models. These techniques require 
large amounts of data to train, and it is difficult to 
design a system that can easily adapt itself or be 
retrained to account for a change in the way the user 
performs gestures. It was therefore decided to use a 
template - based technique as the basis of the 
recognition process. Template based techniques 
compare a time - varying signal to a group of 
predefined templates or of the signal and assess their 
similarity. They are comparable in speed with the 
aforementioned approaches and can be retrained 

RESNA ’94 



Where: is the largest principal coefficient 

PC2 and PC^ are the remaining two 
principal coefficients. 

If a gesture is directional, the PCF will be high since 
PCmax»PC2+PC3. The PCF threshold value that 
distinguishes a directional from a complex gesture is 
user dependent, and is one of the recogniser 
parameters that needs to be set using the calibration 
software. Apart from classifying a gesture as 
directional or complex, PCA is also able to 
determine whether the gesture performed was 
mainly horizontal or vertical thereby further 
reducing the number of candidate gestures. 

lor 

June 17-22, 1994 



93 



HEAD GESTURE RECOGNITION 

This section has described several variables in the 
recogniser that are user dependant. In order to 
enable non*technical people to optimise these 
variables to a particular user, a software application 
called the 'Gesture Recognition Workbench' was 
developed. This application calculates the tremor 
thresholds and facilitates the adjustment of all the 
user - dependent variables, as well as providing the 
facility to create gesture templates. Once the 
recogniser has been calibrated, the variables are 
stored in a file which is automatically loaded ev^ 
time the recognition system is used in an 
application. 

RESULTS 

Testing of the recogniser with six disabled subjects 
is being undertaken. Five subjects have varying 
levels of athetoid cerebral palsy, while one of them 
(D3) has Friedrichs ataxia. The preliminary results 
presented here are based on an eight gesture 
vocabulary, namely the base vocabulary and two 
optional gestures chosen by the subjects. The 
vocabulary was kept small due to the fact that the 
subjects were not trained in performing complex 
gestures, and the test was fairly strenuous. The test 
consisted of the subjects being requested to perform 
each gesture in their vocabulary 10 times over a 
random sequence of gestures. An observer later 
examined video tapes of the sessions in order to 
provide a comparison between the human and 
computer recognition capability. This proved to be a 
useful indication of performance, as some subjects 
performed gestures which were clearly not 
recognisable. 

PRELIMINARY RECOGNITION RESULTS 



Subject 


CRA 


NRA 


D1 


88.8% 


92.5% 


D2 


44.2% 


54.6% 


D3 


87.2% 


97.4% 


D4 


85.0% 


95.0% 


D5 


82.5% 


90.0% 


D6 


98.8% 


98.8% 



CRA = Computer Recognition Accuracy 
NRA = Normal Recognition Accuracy 

The results show that the computer recognition 
accuracy was on average only 6.3% below the 
human recognition accuracy. This figure is 
relatively low considering that a human observer can 
obtain visual clues, such as eye movement or facial 
gestures of the subject, that will aid in identifying 
the head gestures. The computer on the other hand, 
relies solely on the stream of position vectors 
produced by the Polhemus to identify tljp. gestures. 



Both recognition flgures are expected to improve as 
the subjects get more practice at performing head 
gestures. 

FUTURE DEVELOPMENTS 

The disability this research is concentrating on is 
athetoid cerebral palsy. In the near future we expect 
to analyse the movement 'noise' inherent in athetoid 
cerebri palsy, and attempt to extract it from the 
gesture movem^t. This should considerably 
improve the recognition rate. 

Two applications using the gesture recognition 
system are also currently being developed. One is an 
application aimed at young children which teaches 
them aspects of shape recognition. The second, is a 
head gesture assisted direct control system for a 
robotic manipulator!^. 

ACKNOWLEDGEMENTS 

The authors would like to thank the Papworth Group 
and the Addenbrookes Disabilities Services Centre 
for their help during the user trials. 

REFERENCES 

[1] W Harwin (1991), Computer recognition of the 
unconstrained and intentional head gestures of 
physically disabled people, PhD Dissertation at 
Cambridge University. 

[2] J Treviranus (1992), Quartering, Halving, 
Gesturing: Computer Access Using Imprecise 
Pointing, RESNA 92 Conference proceedings, pp 
374 - 376. 

[3] G Hamman, (1990), Two switchless selection 
techniques for using a headpointing device for 
graphical user interfaces, RESNA conference 
proceedings, pp 439-440. 

[4] A I Tew, C J Gray, (1993), A Real Time Gesture 
Recogniser Based on Dynamic Programming, 
Journal of Biomedical Engineering, vol. IS, 
PP181-187. 

[5] A Waibel, K F Lee, (1990), Template Based 
Approaches, Readings in Speech Recognition, pp 
113-114. 

[6] W A Mceachem, C Perricos RD Jackson,(1993), 
Head Gesture Assisted Direct Control of a 
Robotic Manipulator, ICORR 94 Proceedings 

Costi Perricos 

Cambridge University Engineering Dept 
Trumpington Street 
Cambridge CB2 IPZ, 

United Kingdom 

108 

June 17-22, 1994 



94 



RESNA ’94 



The Design and Development of a Computer-Based System for Assessing 
and Training Two-Dimensional Language Representation 

Patrick Demasco, Beth Mineo, John Gray, Rachel Bender 
Applied Science and Engineering Laboratories 
University of Delaware/A.I. duPont Institute 



ABSTRACT 

This paper describes a system that can be used in the 
design and delivery of picture-based instruction for 
AAC users with severe cognitive impairments. The 
system design, which is based on previous research in 
picture understanding, allows interventionists to cre- 
ate images that are abstracted along the dimensions of 
color, size and detail. These images can be embedded 
in instructional protocols used for assessment and 
training. 

BACKGROUND 



into instructional screens and to define scripts that 
control the instructional protocol and data collection. 

DESIGN 

The system design reflects the task requirements of 
image manipulation and protocol development. It is 
comprised of two m^jor components: the Image 
Browser and the Protocol Browser. Combined they 
support the design and delivery of instructional proto- 
cols that can be used to assess or train an individual’s 
picture understanding along the dimensions of size, 
color, and detail. 



There are many individuals who cannot benefit from 
current picture-based instruction and augmentative 
communication interventions because their cognitive 
sifills are not sufficiently advanced to permit under- 
standing of pictured information. (1-4). 

The Mineo study (4) demonstrates that, in order to 
fully assess client strengths and weaknesses in regard 
to various levels of representation, we must be able to 
systematically manipulate the size, color, and quality 
features of the stimuli. A system enable of such stim- 
ulus manipulation would permit a comprehensive and 
precise evaluation of client skills. This in turn would 
ultimately result in the selection of intervention mate- 
rials appropriate to the client’s level of functioning. 

There is evidence that we can advance an individual’s 
ability to understand two-dimensional representations 
through the application of systematic training proce- 
dures (5. 6). Thus, an instructional system that is capa- 
ble of manipulating the size, color, and detail of 
graphic representations could be used to generate 
stimuli, that could be incorporated into a training par- 
adigm based on proven instructional principles. 

RATIONALE 

The evolution of computer-based technology, espe- 
cially in the area of gr^hics quality, has created the 
potential to create such an instructional system. This 
paper discusses progress made in the development and 
evaluation of an Apple Macintosh-based application 
that can be used for assessment and training of picture 
understanding skills (7). It applies a variety of interac- 
tive and automatic image processing functions to pro- 
vide an interventionist with the ability to easily 
manipulate image abstractions. This component is 
combined with the capability to embed these images 

*S 

, ^ I if 



Conceptual Model 

The features size, color and detail can be thought of as 
dimensions in a space of possible picture representa- 
tions (Figure 1). Along each axis, a number of charac- 
teristic pictures can be defined. The origin is described 
as a full size color photographic quality representa- 
tion. Along the size dimension, the representation 
ranges from full-size to miniature. Along the color 
axis, the representation ranges from full-color to gray- 
scale (as in a black and white photographs) represen- 
tation. Detail as we have defined it ranges fi'om photo- 
graphic quality to line drawing, but we have also 
defined an intermediate point that is cartoon-like; that 
is. it consists of a relatively small number of colors 
that are used to create solidly shaded areas of impor- 
tant features. 




Figure 1 . Picture space as a function of three 
dimensions of abstraction. 



109 



RESNA’94 • June 17-22, 1994 



PY AVAILABLE 



95 



Computer-based Picture Instruction 




Figure 2. Incremental progression from photograph to line drawing 



It is important to understand that these axes define a 
continuum of representational possibilities. A specific 
representation can exist anywhere in this space (since 
these features are independent) and not just along the 
axes. In addition, it is possible to specify as fine an 
increment as is needed to advance a user’s under- 
standing. Figure 2 shows a progression from photo- 
graph to line drawing along a sequence of ten 
representations. 

Image Browser 

The Image Browser is responsible for image acquisi- 
tion and image manipulation. Acquisition is supported 
through an import feature that rea^ TIFF files (a com- 
mon interchange format) that are produced from either 
a video-based capture system or scanner. Image 
manipulation is a multiple step process that includes 
image preparation, abstraetion preparation, abstrac- 
tion definition, and abstraction creation, 

Image Preparation = The image preparation phase 
includes the process of masking and cropping. Mask- 
ing removes the image background from the object, 
allowing it to be displayed on the white screen back- 
ground without a trace of its rectangular photogr^hic 
origins. While it is possible to photogr^h an object 
on a solid white background, it is almost impossible to 
totally eliminate subtle color and shading differences 
between the resulting image background and the 
screen background. Cropping is simply the process of 
eliminating any of the background area that is extrane- 
ous to the object, It defines a bounding rectangle 



around the object that is used in subsequent display 
and sizing operations. 

- The abstraction preparation 
phase is necessary to support manipulation along the 
dimension of detail. While size and color abstractions 
can be created automatically from parameters, detail 
requires the creation of two interm^ate images. The 
first is a line-based representation that can be created 
by applying an edge detection algorithm. The second 
representation is used for cartoon-like images and 
consists of a small number of colors. This can be 
obtained through a process called quantization, where 
the image is analyzed for its dominant colors and is 
then remapped using this small set. These two repre- 
sentations ^ong with the original image can then be 
blended to form intermediate representations along 
the detail dimension. 

Abstraction Definition and Creation- For a given pro- 
tocol, the interventionist must decide which abstrac- 
tions we to be manipulated and how finely they should 
be varied. During this process, a series of named 
abstractions are created that define the representations 
to be used in the protocol. They are similar to named 
Styles used in many word processing programs. Dur- 
ing this process the interventionist defines numeric 
parameters for each dimension (size, color and detail) 
through a dialog box. Subsequent to this definition 
process, abstractions can be applied to any of the 
images that were brought into the system. 



110 



RESNA’94 • Jim# 17-22, 1994 



96 



Computer-based Picture Instruction 

Protocol Browser 

The Protocol Browser allows a clinician to define a 
series of instructional screens. Each screen defines an 
arrangement that corresponds to the desired instruc- 
tional task. For example, a matching task might 
involve a full size photogn^)hic representation at the 
top of the screen with four abstract representations at 
the bottom. Each screen can then be instantiated with 
specific images that have been defined in the Image 
Browser, or the system can assign images according to 
a specified randomization criteria. The second pot of 
protocol development involves defining an instruc- 
tional script. The information contained within the 
script includes the duration that each screen is shown, 
inter-stimulus intervals, prompting, and actions to be 
taken based on student input. Currently, scripts must 
be developed using C++, but future versions will 
include a user accessible interpreted scripting lan- 
guage. 

DEVELOPMENT 

The prototype software has been implemented on an 
Apple Macintosh Quadra 900 with a 24 bit graphics 
adapter, video and scanner input, and removable mass 
storage. The application is written in C++ using the 
Apple’s MacApp™ application framework. M^or 
software components include the gr^hics data model, 
user interface components, and imago processing rou- 
tines. 

In the design of the Image Browser, a number of inter- 
esting design solutions were developed that combined 
traditional image processing techniques with interac- 
tive control. For example, tradition quantization auto- 
matically creates reduced color images from a source 
image and a specification of the number of desired 
colors. Unfortunately, most algorithms have not been 
designed to work with a very small number of colors 
(e,g., less than 10). We have developed an interactive 
quantization technique that allows the user to “pick” 
the desired colors directly from the image. The system 
then re-maps the image to the desired color set. This 
has the advantages that color selection is perceptually- 
based and, by virtue of the fact that it is interactive, 
the user can try different selection sets until the 
desired result is obtained. This is not possible with an 
approach that is based solely on automatic quantiza- 
tion. 

EVALUATION 

The current system has undergone an extensive inter- 
nal evaluation. This alpha testing, which has been pri- 
marily limited to the Image Browser, has uncover^ a 
variety of software bugs and problems. For example, 
the masking process was found to be inaccurate with 



certain objects such as a comb. In addition, a number 
of additional features have been suggested through 
this phase. The system is scheduled for field testing in 
the spring of 1994. During this phase a test protocol 
will be evaluated with a number of students who have 
cognitive limitations. 

REFERENCES 

(1) Sevcik, R„ & Romski, M.A. (1986). Representa- 

tional matching skills of persons with severe retar- 
dation. Augmentative and Alternative 

Communication. 2. 160-164, 

(2) Mirenda, P., & Locke. P. (1989). A compailson of 
symbol transparency in nonspeaking persons with 
intellectual disabilities, Jpurnaj gf Speq ph 
Hearing Disorders. M. 1 3 L 140. 

(3) Mizuko, M., & Reichle, J. (1989). Transparency 
and recall of symbols among intellectually handi- 
capped adults. Journal of Speech and Hearing Dis- 
orders. 627-633. 

(4) Mineo. B.A. (19901. A Feature-based Approach (p 
the Evaluation of Representation^ Capabilities. 
Paper presented at the Fourth Biennial internation- 
al ISAAC Conference, Stockholm, Sweden. 

(5) Dixon, L. (1981). A functional analysis of photo- 
object matching skills of severely retarded adoles- 
cents. Journal of Applied Behav ior Analysis. H. 
465-478. 

(6) Shane, H., & Blau, A. (1981). Vocabulary selec- 
tion in nonspeech communication. Paper present- 
ed at the meeting of the Pennsylvania Speech and 
Hearing Association. Philadelphia, PA. 

(7) Dcmasco, P.W, & Mineo, B.A. (1992). Enhancing 
Picture-Based Communication; Assessinent and 
|njii(ruptioit, Paper presented at the Fifth Biennial 
International ISAAC Conference, Philadelphia, 
PA. 

ACKNOWLEDGEMENTS 
This work has been supported by the National Insti- 
tute on Disability and Rehabilitation Research 
(#H133Q 10189), Additional support has been pro- 
vided by the Nemours Foundation, 

Patrick Demasco 

Applied Science and Engineering Laboratories 
A, I. duPont Institute 
1600 Rockland Road. P.O. Box 269 
Wilmington, Delaware 19899 USA 
Internet; demasco@asel.udel.edu 



111 

RESNA’94 • J«ne 17-22, 1994 



97 



THE SPEECH TRANSLATOR - 

A COMMUNICATION AID FOR SEVERELY IMPAIRED SPEAKERS 

Andrew R Gammie BA 

Bath Institute of Medical Engineering, United Kingdom 



ABSTRACT 

The aim of the project is to produce a unit that 
recopises the ^storted utterances of a severely 
impaired speaker, which are often very difficult for 
the listener to understand. The translator then 
speaks a word or phrase with a speech synthesiser 
that the listener can understand. The device has 
therefore been called a ’speech cleaner’ by some. 
We have bad field trials with a number of potential 
users. 

BACKGROUND 

Development of a translator communication aid for 
severely impaired speakers has been under way for 
two years. The concept of recognition of non- 
verbal utterances has been proposed before (1,2), 
but people with inconsistent vocalisations, eg some 
cerebral palsied individuals, have not been able to 
use the devices. New technology presented here 
allows them to do so. 




DESIGN 
a) Software 

The recognition technology used has been 
developed at the Royal Military College of Science 
(RMCS), Cranfield University, Shrivenbam, UK 
(3,4). Tire time encoded speech (TES) method 
used functions well in noisy environments. In 
addition, and more importantly, the method can 
recognise the repeated, elonpted or stuttered 
utterances typical of many disabled speakers. 

We have developed the user interface for a TES 
recognition unit using a single switch input and a 
small liquid crystal display. The programme flow is 
illustrated in figure 1. Training of the translator is 
done by the user repeating the word five times, and 
at present storage of ten distinct utterances are 
possible. The phrases and words stored in the unit 
can be edited by connection of a computer terminal 
or by the LCD display. 



Figure 1. Programme flow 



bl Hardware 

The system is based around a Bowman 68010 
microprocessor board (SDS Ltd, High Wycombe, 
UK), with 1 Mb of ROM/RAM memory space. 
Custom boards built by RMCS Shrivenham 
perform the analogue signal processing and codmg. 
Figure 2 shows a block diagram of the system. 

EVALUATION 

Clinical trials have been carried out with 15 users 
and the feedback has been very encouraging. 

The system has throughout demonstrated extremely 
well its ability to recognise stuttered or elongated 
speech inputs. Various distortions have been 
imitated, with performance results being the same 
as with normal speech. 



112 



RESNA ’94 • June 17-22, 1994 



98 



The Speech Translator 




Figure 2. 

System hardware block diagram. 



of the unit should also be questioned due to the 
level of motor control required for simultaneous 
speech and switch use. This could be done by use 
of a sound operated switch being sensitive to the 
presence of an input, but the practicalities are 
difficult and require detailed consideration. These 
and other developments would enable formally 
assessed trials to take place. 

This further work is currently under consideration. 
The technology is undoubtedly a major advance for 
severely impaired speakers, enabling them to have 
access to some of the enormous benefits speech 
recognition technology can bring. 

ACKNOWLEDGEMENTS 

We are very grateful to the Nuffield Foundation for 
their generous financial support of this project, and 
to Domain Dynamics Ltd for making the TES 
technology available to us. 



ERIC 



Enthusiasm for these demonstrations of the concept 
has been high amongst potential users and 
therapists alike. RMCS Shrivenham have 
demonstrated recognition accuracies of 96%, even 
in the presence of high background noise. 

Clients have been taken from a number of groups 
of disability, on the suggestion of the therapists 
involved who saw applications for several different 
groups. These included cerebral palsied children 
and adults (4), Downs syndrome children and 
adults (2), adults with learning difficulties (3), 
adults with compound disabilities including 
prof ound hearing loss (3) and adults with severe 
head injiuy (3). Other groups suggested have been 
people with Parkinson’s disease, strokes and 
laryngectomies. 

During trials, clients have suggested a number of 
areas of use for the device. These ranged from 
communication in the home and institution to 
shopping, cinema and other leisure uses, given that 
the translator would eventually be more portable. 

DISCUSSION 

Further development of the device is planned. This 
will include a new speech synthesiser, development 
of the user interface arising from clinical trials 
results and miniaturisation of the unit in order to 
make it portable. The use of a switch for operation 

RESNA’94 • 



References 

1. "Speech recognition technology for individuals 
with disabilities" 

Noyes JM, Frankish CR, Augmentative and 
Alternative Communication 92 (8) 297-303. 

2. "Preliminary studies for a simple personal 
computer interface using voice/sound recogmtion 
to facilitate communication.." 

Boonzaier DA, Limon A, Proc RESNA 14th 
Annual Conference, p.176-178. 

3. "An automatic diagnostic routine for the 
selection of acoustic utterances produced by 
severely impaired speakers" 

Warner AG, Hughes RD & King RA, Proc 7th 
FASE Symposium, Speech ’88, Edinburgh UK, 
August 1988. 

4. "A direct voice input man-machine interface 
strategy to provide voice access for severely 
impaired speakers" 

Warner AG, Hughes RD & King RA, Proc UK IT 
Conference, Southampton UK, March 1990. 



Andrew Gammie, 

Electronics Design Engineer, 

Bath Institute of Medical Engineering, 
Wolfson Centre, 

Royal United Hospital, 

Bath BAl 3NG 
United Kingdom 

113 



June 17-22, 1994 



99 



TALKSBAC: A PREDICTIVE ADAPTIVE CONVERSATION SYSTEM 
FOR NON-FLUENT DYSPHASICS. 

Annalu Waller^, David Millar^, Fiona Dennis^»^, Alistair Y. Cairns^, Peter Gregor^, 

Janet K. Brodie^, Alan F. Newell^ Ketron Morrison^ 

^ MicioCentre, University of Dundee, Scotland, ^ Dundee Speech & Language Therapy Service 



Abstract 

A communication system, called TalksBac, has 
been developed which predicts phrases, 
sentences and story texts for conversational use. 
The system has been designed to augment the 
communication of adults with non-fluent 
dysphasia. The retrieval system uses a dynamic 
semantic network to identify probable 
conversational items which relate to user- 
selected parameters. 

Background 

Dysphasia following stroke leaves people with 
expressive and/or receptive difficulties of 
varying degrees, affecting both spoken and 
written communication. Generally, traditional 
AAC intervention with this client group has been 
unsuccessful due to the high levels of linguistic 
ability required to use currently available 
systems [1]. 

Although word prediction is beneflcial for people 
with learning, spelling and language problems, 
most phrase and sentence retrieval systems 
remain code-based [2]. Preliminary work with 
two dysphasic clients using a predictive 
communication system indicated that some 
clients can recognise and choose famUiar words 
and/or phrases within a conversational setting [3, 
4]. A novel story telling facility aUows for the 
retrieval and narration of longer chunks of text 
[4,5]. 

System Overview 

TalksBac (Talking And Learning Knowledge 
System for Better Aphasic Communication), a 
computer based predictive communication 
system, has been designed to guide the dysphasic 
user through a series of parameter choices until a 
target sentence or story title is recognised. 

The TalksBac software is designed to run on a 
portable version of the Apple Macintosh 
computer. The internal Speech Manager provides 
access to speech synthesis without the need for 
additional external hardware. 

The software reduces the complexity of 
information retrieval by restricting ie number of 
choices presented to the user at any one time. 
First the user is asked to select the person to 



whom they are talking from a list of probable 
names. The program then offers a list of topic 
items associated with the chosen person. 
Predicted sentences and story titles related to the 
current person and topic/s are presented for 
selection . 

The number of parameter choices and 
conversational items can vary according to 
individual users e.g. predictions can be presented 
one at a time or in a list of up to four. Selected 
sentences are spoken inunet^tely whereas the 
selection of a narrative results in the narrative 
text being displayed with an appropriate control 
panel . This control panel allows the user to 
control the rate of the story narration. 

System Design 

The system has three main aspects (fig 1): the 
user interface, the prediction engine which 
controls a semantic network and database, and 
the carer interface. 




Fig 1: The System 
The User Interface 

The interface is divided into the user interface 
and the carer interface. The user interface allows 
the user to retrieve and "speak" pre-stored 
information, while the carer interface controls 
the input and management of the conversational 
data. 

The user interacts with the interface by selecting 
labelled buttons with a built-in track ball. The 
number of buttons is minimised so that only the 



ERIC 



100 



RESNA ’94 



114 

• June 17-22, 1994 



TALKSBAC 



most probable options are offered. The options 
(i.e. the labels on the buttons) are compiled by 
the prediction engine. 

Conventional storage and retrieval systems use 
static codes to access prestored information 
whereas the TalksBac system is adaptive. 
Choices made by the user activate the Prediction 
Engine which continually adapts a semantic 
network to reflect the user’s conversational 
patterns. This reduces the need for users to 
remember what has been stored and where it is 
stored in the system. A specific conversational 
item may be accessed along a number of 
pathways. 

The Prediction Engine 

The prediction engine uses a semantic network to 
determine which topics/conversational items 
should be offered to the user. Conversational 
items in the database (phrases, sentences and 
story titles) are linked to one or more topic 
words. Each discrete topic word is represented in 
the network as a node (e.g. Canada and skiing in 
fig. 2). People are treated as topic words, but are 
tagged as being people. Topic nodes are 
connected alphabetically for direct retrieval. 
Nodes are also connected by weighted links 
which indicate the degree of association between 
topic words (e.g. 0.5 between Canada and 
skiing). These weights change dynamically to 
reflect use of the system. In addition to the 
weighted links, each topic node has a list of 
database items (sentences and story titles) which 
have been associated with the topic word (e.g. 
numbers 1, 2 and 3 in fig. 2 represent items in 
the database). 



a 




z 

Fig. 2: Part of the semantic network. 



The choice of topic parameters selected by the 
user has a two-fold purpose. First, it provides a 
template from which a list of probable items can 
be predicted; and second, the links between topic 
nodes are modified to reflect the user’s choice. 
The prediction engine thus dynamically alters 
the form of the semantic network. This dynamic 
adaptation not only reacts to topic ’’closeness”, 
but also reflects frequency and recency of use. 

The Carer Interface 

Entering new conversational infonnation into the 
system is done via the carer interface due to the 
dysphasic users' expressive language difficulties. 
Tbe carer can enter sentences, stories and 
people’s names by simply typing in the required 
text at the appropriate prompts. The system 
parses sentences and story titles, ignoring all 
function words. The remaining words are used as 
associated topic words and the carer is given the 
opportunity to add or delete words from this list 
The carer can also add associated words to 
people. A facility to delete or modify 
information is available. 

Application and Evaluation 

A small group of non-fluent dysphasic clients 
has been selected to take part in single case 
studies to evaluate the use of TalksBac as a 
conversation aid. A battery of formal tests has 
been administered to each client and wUl be used 
as baseline data. The tests will be repeated at the 
end of the project and will be used to indicate 
any change in the clients' comprehension and 
communicative abilities. 

The first year of this two-year project has been 
spent developing the software. The second year 
will involve client use of individual systems. 
Clients and their carers have been trained to use 
the device and each client has been given a 
system to use for a year An analysis program 
allows the evaluation of prediction algorithm as 
well as identifying how often and with whom the 
system is used. 

Acknowledgements 

Joint funding for this two year project has been 
received from the Scottish Home and Health 
Department, the Leverhume Trust, the Esmee 
Fairbaim Charitable Trust and Hydro Electric. 

References 

[1] Blackstone, S. (1991). Persons with severe 
aphasia: What does AAC have to offer? 
Augmentative Conununication News, 4:1, 
1-3. 



115 

RESNA’94 • Jane 17-22, 1994 



101 



TALKSBAC 



[2] Newell, AP., Amott, Ji., Booth L., Beattie, 

W., Brophy, B. (1992). Effect of the "PAL" 
word pr^ction system on the quality and 
quantity of text generation. Augmentative 
and Alternative Communication, 8, 304- 
311. 

[3] Bioumley, L., Cairns, A.Y., Amott Ji. 

(1992). Evaluation of a personalised 
conununication system developed with 
phasic adults. Augmentative and 
Alternative Communication, 8:118. 

[4] Waller, A. (1992). Providing Narratives in an 

Augmentative Communication System. 
PhD. Diesis, University of Dundee, 
Scotland. 

[5] Waller, A., Broumley, L., Newell, A.F., Aim, 

N. (1991) Predictive retrieval of 
conversational narratives in an 
augmentative communication system. 
Proceedings of the 14th Annual RESNA 
Conference, Kansas City, 107-108. 



Contact Address 

Annalu Waller Ph.D. 

MicroCentre 

Dept, of Mathematics & Computer Science 
Die University 
Dundee DDl 4HN 
Scotland 

Tel: (0382) 344753 

Fax: (0382) 23435 



Er|c 102 



RESNA ’94 » Uune i: 



-22, 1994 



EXTENDING THE FUNCTIONAL DAY FOR POWERED WHEELCHAIR BASED TECHNOLOGIES 

Ed Snell, GET., Lanry Silver 

Bloon/iew Childrens Hospital 
Willowdale, Ontario, CANADA 



ABSTRACT 

Technologies are increasingly being added to 
powered wheelchair platforms incorporating 
considerations for augmentative communication, 
writing aids (laptop computers) and environmental 
control. The self-contained rechargeable batteries 
in most add-on technologies will not provide a 
days operation for the more active client. 
Promotion of client independence and quality of 
life necessitates operation of assistive 
technologies for the full functional day. An efficient 
switching power supply was designed to convert 
the 24 volt wheelchair batteries into three 
separately adjustable DC output voltages to power 
wheelchair based technologies. 



INTRODUCTION 

Clients with more severe and complex disabilities 
are being integrated into school systems and 
communities, putting greater demands on the 
powered wheelchair platform. Wheelchair 
manufacturers have responded to this increasing 
demand by putting more sophisticated electronic 
control systems in place thereby facilitating client 
customization. Some manufacturers have provided 
features on their control systems in a generic 
manner that allows the client to use the driving 
control to perform other peripheral functions as 
.nay be necessary (ie. tiiters, recliners). 

Personal communication (whether face-to-face, 
voice or written output) and environmental control 
are examples of client needs frequently 
encountered in combination with powered 
wheelchairs and seating. Voice output devices 
have an operating time of 8-12 hours on a single 
charge, depending on the device. This would 
provide the user with operation over a school day, 
but not over a complete waking day. Laptop or 
notebook computers are used for all forms of 
personal communication in conjunction with the 
appropriate software and access method. These 
computers however only have sufficient battery life 



to operate for approximately 2-3 hours, depending 
on the model. This is not enough operating time 
for practical use in a school setting and makes the 
user dependent on others (and their skill level) to 
connect chargers during the day. The recharging 
process either requires the technology to be 
temporarily removed from the wheelchair or that 
the user be confined within a wire’s length of the 
electrical outlet. This problem is compounded if 
more than one technology is mounted on the 
wheelchair and each requires charging. 

The powered wheelchair usually has two 12 volt, 
55 ampere-hour lead acid batteries connected in 
series to supply 24 volts to the control system and 
motors. Only extremely active wheelchair users 
require the full battery capacity to operate the 
wheelchair for a full waking day and 
manufacturers offer optionally larger battery 
capacities. Notebook conrrputers (which are quite 
power hungry in comparison to AAC devices) 
require approximately 1% of the wheelchair's 
battery capacity per hour for full operation. That 
portion of the battery capacity not requited for 
daily wheelchair operation can be converted into 
voltages to power other devices. Charging only the 
wheelchair batteries overnight will provide 
sufficient power to operate the wheelchair and all 
peripheral devices for the full day. An efficient and 
comprehensive method of power conversion was 
required to facilitate this. 



DESIGN CRITERIA 

Some computer notebook manufacturers have 
developed adaptors which allow the computer to 
be connected to the automotive cigarette lighter 
outlet. The outlet provides 12 volts DC and the 
adaptor converts this into the necessary voltages 
for the computer. Some rehabilitation centres use 
these adaptors on powered wheelchairs by 
connecting them to one of the two 12 volt 
batteries. This indeed does work but the uneven 
loading of the two batteries will considerably 
shorten the battery life. A wheelchair voltage 



117 

RESNA’94 • June 17-22, 1994 



103 



EXTENDING THE FUNCTIONAL DAY 

converter should connect across both batteries 
and be protected from accidental connection in 
reverse polarity. 

For the voltage converter to work with a number 
of different devices the outputs must be 
adjustable. If connection is being made to 
rechargeable batteries for extended periods of 
time, the voltage adjustment must be continuously 
variable versus step adjustable. These voltages 
are quite critical to prevent overcharging while 
maintaining a reasonably full charge on the battery 
pack. It should be noted that nickel cadmium 
batteries will develop a charge memory if 
maintained on a float charge. These batteries 
should be discharged on a regular basis (ie. once 
a week) to retain their capacity. 

Anyone who has worked around wheelchairs will 



protection for electronic components and does not 
reset itself when the overload is removed. The 
voltage converter needs to be short circuit and 
overload protected in a manner that will allow 
operation to resume when the problem condition 
has been removed. 

Severe transients are produced across the battery 
terminals when the wheelchair is driven. Protection 
from these transients needs to be provided for the 
voltage converter and any devices it powers. 



FEATURES 

The Bloorview Battery Adaptor is a switching 
power supply which has three independent output 
channels with a common ground. The output 
ground is also common to the most negative 




know that wires get pinched, cut and the plugs 
pulled off cables. The wheelchair batteries have 
sufficient power to cause personal injury and 
damage to electronics that are not short circuit 
protected. A fuse does not provided adequate 



terminal of the wheelchair battery series 
combination. Input power to the adaptor is 
supplied directly from the 24 volt wheelchair 
battery supply via a battery compartment fuse and 
power connector assembly. The adaptor is 



- 118 

RESNA ’94 • June 17-22, 1994 BEST COPY AVAILABLE 



104 




EXTENDING THE FUNCTIONAL DAY 



instantaneously protected from application of 
power in reverse polarity and will resume 
operation when the polarity has been corrected. 

Each adaptor output is independently and 
continuously adjustable from 0 to 18 volts through 
a single turn potentiometer on the internal circuit 
board. Output connections are provided through a 
female 9-pin "D" panel connector for custom 
cabling and independent outputs are provided on 
three power connectors. Output regulation is 0.5% 
of output plus 0.1 volts. The standby input current 
is 40mA at 24 volts. 

Each of the outputs are electronically short circuit 
protected and will resume normal operation when 
the shorting condition is removed. The power at 
each output is limited to 7 watts but outputs can 
be paralleled for higher power devices. Lower 
voltage devices can be supplied larger currents 
than can higher voltage devices according to, 
POWER = VOLTAGE x CURRENT. 

Output transient protection is provided through an 
LC pi filter and computer grade switching 
capacitors. Input power is protected by an LC pi 
filter and transient protection circuitry. 

The internal printed circuit board is self-contained 
and can be removed quickly from the chassis for 
senricing or board swapping. 



clients running out of wheelchair battery power 
after the installation of the adaptor and computer 
system. Non-speaking clients can now 
communicate from breakfast through to the last 
snack at night with power to spare. 



ACKNOWLEDGEMENT 

The authors wish to thank the Bloonriew 
Foundation for their continuing support of the 
Responsive Development Team. 

For further information on this and other Bloonriew 
products contact: 

Mr Ed Snell, CET., 

Coordinator, Rehabilitation Technology, 

Bloorview Childrens Hospital, 

25 Buchan Ct., 

Willowdale, Ontario, M2J 4S9. 

CANADA 



DISCUSSION 



The Bloonriew Battery Adaptors have been 
supplied to several clients mostly for powering 
notebook computers and voice output devices. 
Some computers require multiple voltages and 
some require considerable power. Both of these 
factors reduce the number of other devices that 
can be operated since there are only three outputs 
and each output is limited to 7 watts. 
Environmental control applications included 
powering wireless transmitters such as those for 
powered door openers and remote access. Clients 
who require these transmitters are generally 
unable to independently change the replaceable 
batteries and discover that the battery is dead 
when the door no longer opens (this hopefully 
happens when the client is on the home side of 
the door). 




The first Bloonriew Battery Adaptor was installed 
2 years ago. There have been no reports of 



119 



RESNA ’94 • June 17-22, 1994 



105 



A MORPHOLOGICALLY-BASED WORD PREDICTOR FOR SWEDISH 



Sheri Hunnicutt, Johan Bertenstam and Parimala Raghavendra 
Royal Institute of Technology, Stockholm, Sweden 



ABSTRACT 



In order to address the problems of undesired 
morphological forms in prediction lists, limited 
space available for lexicons, and limited space for 
predictions, a morphologically-based word 
predictor and accompanying base form lexicon 
have been developed for Swedish. Nouns, 
adjectives and verbs, which have multiple forms in 
Swedish, are chosen in a two-step process, by first 
choosing a base form, and then choosing the 
desired inflected form fi’om an automatically 
generated list. An evaluation has been planned 
which will explore the possible benefits of this 
program for users with reading and writing 
difficulties by comparing texts written with and 
without this facility. Several unresolved questions 
regarding implementation and evaluation are 
discussed. 



BACKGROUND 

Several word prediction programs have been 
developed for Swedish and have been in use for a 
number of years [1,2,3]. The major word lists 
which accompany these programs have 
traditionally been fi^equency-ordered. The most 
recent evaluation of the Swedish word-prediction 
program Profet (English: Prophet), has shown 
improvements for some users in the areas of 
spelling, syntax, length of time in use before tiring 
and comprehension [4]. 

STATEMENT OF THE PROBLEM 

An observation that has been made by program 
evaluators is that users can become frustrated 
when offered one or more undesired forms of a 
desired word in a prediction list. This situation 
occurs whenever more fiequent forms exist, since 
the first few letters of the words in a noun or 
adjective declension or a verb conjugation are 
usually identical. Although word frequency is a 
major and powerful factor in successful word 
prediction, an informal test of the occurrence of 
less frequent forms provided an interesting result: 
35% of all declinable/conjugable words in a 
chosen text did not appear in the first 5-element 
prediction list that contained a form of the word. 

RESNA ’94 



A companion problem is that in a frequency- 
ranked lexicon of a given size, all forms of all 
words do not appear, so that a desired form may 
not even appear in a later prediction list. This is 
particularly evident in a language such as Swedish 
in which a verb conjugation contains 20 forms (if 
you include the passive forms), an adjective 
declension, 7 forms and a noun declension, 8 
forms. Including all forms in a lexicon would 
drastically reduce the semantic coverage through 
reduction in the number of unique root morphs. 

Another possible constraint is the necessity to 
limit the number of entries in a prediction list. 
This constraint may occur because of visual or 
cognitive limitations, or because of limits on 
presentation area. 

APPROACH 

The solution which has been adopted in a version 
of Profet called Profend is to create a lexicon 
containing only base forms (verb infinitive, utrum 
nondeterminate adjective and nondeterminate, 
nonpossessive singular noun), and to develop an 
algorithm which allows the user to choose these 
content words in a two-step process. This 
approach presents the user with a list of words in 
consistent morphological form, and allows a given 
size lexicon to achieve significantly greater 
coverage. 

IMPLEMENTATION 

A lexicon was first constructed which is a 
combination of the 2,000 most fiequent Swedish 
words [5] and a list of about 7,000 words in their 
base forms which Swedish students are expected to 
know upon graduation [6]. Any of the 2,000 most 
fiequent words which did not occur in their base 
forms were then replaced by the base form. This 
lexicon was subsequently marked with codes 
corresponding to morphological classes (5 for 
verbs, 6 for adjectives and 11 for nouns). This 
categorization used the set of codes utilized in the 
Chalmers Lexicon [7]. 




June 17-22, 1994 



106 



A Morphologically-based Word predictor 



A table was then constructed which contains 
entries consisting of a categoiy code, a 
subcategorization code and a list of word endings. 
The subcategorization codes enumerated in the 
Chalmers Lexicon, are the actual letter strings 
which terminate a subset of words in the 
morphological categoiy. To produce a desired 
morphological form, the algorithm accesses the 
base form, deletes the subcategorization code letter 
string from the base form and then afrixes the 
desired word ending. 

In the example below, the first line contains the 
base form for the word "flicka" (girl), the second 
line contains the categoiy code "si" (noun, 
categoiy 1) and the subcategorization code "a." 
The third line first lists the inflections for 
indeterminate singular, determinate singular, 
indeterminate pliual, and determinate plural, and 
then the inflections for the associated possessive 
forms. The indeterminate plural form, for 
example, would be constructed by removing the 
subcategorization code "a" from the base form to 
get the stem "flick" and then adding "or" to obtain 
"flickor." 

Baseform: flicka 
si a 

a an or oma as ans ors omas 

It was necessary to provide a separate listing (file) 
of irregular verbs. Since the forms of many of 
these words differ fr^m one another to such a great 
extent, this file contains whole words. An example 
is the verb "gOra" (to do, to make). The list begins 
as follows: infinitive, imperative, present 

participle, present, simple past, past participle and 
verbal adjective (neutrum), verbal adjective 
(utrum),... 

gSra gSr gSrande gSr gjorde gjort gjord gjort.. 

When a user first sees the base form of a desired 
inflected word in a prediction list and chooses it, 
the program retrieves the word's morphological 
categoiy from the main lexicon. Then, using the 
information in the table described above, the 
program generates all forms of the word and 
presents the new list to the user. Thus, the desired 
word is chosen in a two-step process. Of course, 
only one step is necessary if the word only has one 
form. 

After a word has been completely specified and 
entered in the text, its base form is entered in the 
"subject lexicon" which contains words recently 



RESNA ’94 



used as well as words used when previously 
writing about the current topic. 

EVALUATION 

A procedure has been designed in which the two 
programs Profet and Profend will be compared to 
determine user preference and potential benefit in 
writing tasks. The evaluation will be carried out 
with participation of two subjects who have 
reading and writing difficulties (with no motoric 
complication). A single-subject AB 1 AB 2 design 
will be employed with replication across subjects 
[8]. The order of introduction of the programs (Bj, 
Bj ) will be counterbalanced. 

Baseline data of the subjects' writing abilities will 
be collected by asking them to copy a given text 
and to write a free text. They will be trained on 
Profet or Profend first, depending on the condition 
that they are assigned to. The program will be 
demonstrated and one of the investigators will 
work with the subjects until th^ can use the 
program independently. Subjects will be asked to 
use the program for 3 to 4 weeks, at the end of 
which th^ will be administered a probe test. After 
a break of one week, thq^ will learn to use the 
other program and the above procedure will be 
followed. The probe tests will consist of copying, 
taking a dictation and writing free texts using 
Profet and Profend. 

The texts will be analysed for morphological (and 
other) errors and ease of comprehension. The 
subjects will also be interviewed after each of the 
probe tests concerning there percieved effectivness 
and ease of use of the programs, and after both 
probe tests, concerning their preference. 

DISCUSSION 

There are several questions regarding 
implementation of Profend which are, as yet, 
umesolved. One question is concerned with 
whether all forms of a word should be presented. 
Verb passives, for example, are formed from the 
corresponding active form by affixing an "s." This 
addition could be provided for by showing a 
prediction list containing the single entry "s" after 
any verb form is chosen. No extra k^stroke would 
be necessary if the passive form was not desired. 
Also, some conjugations and declensions have 
identical spellings for two forms. On the one 
hand, it seems urmecessary to list the same 
spelling twice in one prediction list, but on the 
other hand, it seems that listing forms in a 

121 

• June 17-22, 1994 



107 



A Morphologically-based Word predictor 

consistent order would mini mize the cognitive 
load of making the correct choice. 

It would not be surprising if users foimd it 
coimterintuitive to access extremely irregular verb 
forms firom the verb's base form. Looking at the 
example of an irregular verb given above, one 
notes that it might be necessary to type "gfl" to 
access the past participle "gjort." An alternative 
is to include aU forms of irregular verbs in the 
main lexicon. However, one would want to 
present them to the user in a prediction list which 
contrasts in appearance or placement sufficiently 
from the normal prediction list so as not to lessen 
the advantage of otherwise consistent presentation. 

Additionally, a procedure will need to be 
developed to allow users to specify a 
morphological category for new words before 
being assimilated into the main lexicon. Word 
class could be specified by choosing a set of words 
judged to be similar. After choice of word class, 
using the program's generation facility, it would be 
possible to present possible conjugations or 
declensions to the user for approval. 

The need to limit presentation area in the switch 
access program SAW [9] has led to a request for 
our co-operation in providing a word-ending 
paradigm for the Swedish version of this system. 
With the lexicon and algorithm described above, 
the user, via a short list of baseform words, will be 
able to access all their morphological forms. 

The results of the evaluation will probably depend 
on the specific reading and writing problems of 
the users. It is predicted that the two-stage word 
choice procedure provided in Profend, in which a 
first prediction list shows words which differ 
substantially firom each other and a second 
prediction list highlights different grammatical 
forms of the same word, will aid some users in 
making morphologically correct choices. On the 
other hand, some users may not have the formal 
linguistic knowledge to be able to choose an 
inflected word using this two-step process. 
Linguistic competence will be particularly 
important if users are to choose inflected forms 
firom base forms which differ radically in spelling. 
Recognition of related base forms may, however, 
be aided by the synthetic speech facility. 



ACKNOWLEDGEMENT 

This research has been supported by the Swedish 

National Board for Industrial and Technical 

Development (NUTEK). 

REFERENCES 

[1] Hunnicutt, S. (1986): "Lexical Prediction for a 
Text-to-Speech System," in Communication 
and Handicap: Aspects of Psychological 
Compensation and Technical Aids, E. 
Hjelmquist & L.-G. Nilsson, eds., Elsevier 
Science Publishers. 

[2] Carlson, R., Granstrdm, B. and Huimicutt, S. 

( 1 990) : "Multilingual Text-to-Speech 
Development and Applications," in Ainsworth, 
W.A., ed. Advances in Speech, Hearing and 
Language Processing, JAI Press Ltd., London, 
England. 

[3] Hunnicutt, S. (1989): "ACCESS. A Lexical 
Access Program," Proc RESNA, New Orleans, 
pp. 284-285. 

[4] Magnusson, T. & Hunnicutt, S. (1992): 
"Swedish Word Prediction: A Follow-Up 
Investigation," Augmentative and Alternative 
Communication, 8/2, Proc. ISAAC conference, 
Philadelphia, p. 151. 

[5] A116n, S. (Ed.)(1970). Nusvensk 
frekvensordbok. (Frequency Dictionary of 
Present-Day Swedish). Vol I. Almqvist & 
Wiksell, Stockholm. 

[6] FrOroth, R (1991): Fdrsta ordlistan. Almqvist 
& Wiksell, Stockholm. 

[7] Hedelin, P. & Huber, S. (1991): "A new 
Dictionary of Swedish Pronunciation" Proc. 
Scandinavian Conference of Computational 
Linguistics, Bergen, pp. 105-1 17. 

[8] McReynolds, L. & Kearns, K. (1983): Single 
Subject Experimental Design in 
Communicative Disorders, Pro-ed., Austin, 
Texas, USA. 

[9] Head, P., Poon, P., Morton, C., Colven, D., 
Lysley, A. (1993): "Switch access to Windows 
3 (SAW) - a new concept in emulation 
techniques", Proc. of ECART 2, May 26-28 
1993, Stockholm, paper 22.2 

Sheri Hurmicutt 

Dept of Speech Conununication and Music 

Acoustics 

Royal Institute of Technology (KTH) 

Box 70014 

S- 100 44 Stockholm, Sweden 



- .122 

RESNA ’94 • June 17-22, 1994 



108 



DEVELOPMENT, PROGRAMMING AND CONCEALMENT OF PRIVATE MESSAGES IN AAC 

SYSTEMS 

Lynn A. Sweeney, Central Michigan University 
Clifford A. Kushler, Prentke Romich Co. 



Abstract 

People who rely on AAC need to express a full 
range of message types. The content and intended 
use of some messages is private and personal in 
nature. This paper describes an initial set of 
guidelines and progranuning strategies for working 
with people who use AAC to effectively develop 
and access messages of a private nature. 

Background 

Many individuals who rely on augmentative and 
alternative communication (AAC) do not program 
or arrange their own messages within AAC 
systems. The AAC user may be prohibited from 
doing so based upon barriers related to age, 
physical, cognitive, linguistic or literacy status, 
and/or lack of opportunity or related control exerted 
by others. As a result, messages of primary 
importance to the individual or specific preferred 
message wording may be overlooked by the 
assisting message/system programmer. The more 
personal, private, or intimate the nature of the 
message or sentiment, the greater the likelihood that 
it will not become part of the AAC system when an 
intermediary message constructor is necessary. The 
message constructor may feel intrusive, 
embarrassed, or reluctant to address the content 
inherent in the private messages of individuals who 
rely on AAC. Yet, these very messages may be of 
the utmost importance to the user in establishing 
and maintaining relationships with others and 
asserting individuality. 

Banajee, Sands, and Schwery [1] postulate that 
communication devices "... are rejected secondary 
to their affective limitations." and note that "... 
constriction of affective expression is not only 
frustrating but limits the potential of the user to 
develop affectively attuned relationships . " Working 
with AAC users to develop messages for 
meaningful communication of afreet and a more 
sensitive approach to the process of progranuning 
were recommended by the authors. They conclude 
that acceptance and use of AAC devices is 
"...facilitated as affective attunement develops 
because the device is utilized for not only 



communicating for general purpose feeling and 
information but also for one’s most private and 
significant feelings." 

Beukelman and Mirenda [2] suggest that "... the 
ultimate goal of an AAC intervention is not to find 
a tedinological solution to the communication 
problem, but to enable the individual to efficiently 
and effectively engage in a variety of interactions. " 
Four purposes are fulfilled in communication 
interactions according to Light [3] . These 
purposes are: 1) communication of needs/wants, 2) 
information transfer, 3) social closeness, and 4) 
social etiquette. Private messages may be required 
within any or all of these types of interactions but 
are weighted more heavily within the first three 
designations. Researchers have noted that 
relatively little attention has been devoted to 
strategies to enhance information transfer and 
social closeness interactions among people who use 
AAC [3,2]. Yet, as noted above, these types of 
interactions (especially social closeness) may be 
most inq)ortant to AAC users and their significant 
others. 

The precision of message content, facility of 
retrieval, and assurance of confidentiality are 
critical to construction and utilization of private 
messages. The goal of the assistive programmer 
then should not be to sensor, monitor, or repress 
such messages but rather to work with the AAC 
user (and others as jq)propriate) to develop 
effective strategies to support communication 
interactions of a private nature. 

OBJECTIVE 

In addition to daily messages intended for more 
public use, AAC users should be provided means 
to construct and/or program private messages 
which may be intended for use in very specific or 
limited circumstances but nevertheless be of critical 
inq)ortance. Independence for message 
programming and arrangement should be a priority 
goal in order to ensure that control and 
prioritization of messages will remain in the 
domain of the person who relies on AAC. 
However, in those situations where this not 



123 

RESNA ’94 • June 17-22, 1994 



109 



PRIVATE MESSAGES IN AAC 



possible on a short or long term basis, the support 
staff/significant others charged with the 
responsibility of progr amming vocabulary or 
constructing communication boards should: a) 
involve the AAC user in the selection, arrangement 
and content of messages to the greatest extent 
possible, b) present opportunities for the AAC user 
to address/select private messages for use in one or 
more communication forms, c) assure that the AAC 
user is able to select/appoint the person/s they 
would like to assist in progr amming private 
messages, d) assure that the methods for concealing 
and accessing private messages are acceptable to the 
AAC user, and e) keep all information regarding 
private messages confidential and inform others to 
support conscientious actions in relation to the gamo 
[4]. 

The objective of this paper is to provide strategies 
for identifying and incorporating private messages 
in AAC systems relative to the above guidelines 
and based upon clinical experiences. 

Method/ Approach 

Private messages and conversations serve a variety 
of functions for everyone. Some of those functions 
include: a) development of personality, autonomy, 
and independence, b) establishing and maintaining 
a personal life and lifestyle, c) seeking direction, 
comfort, and support, d) transcending fear, grief, 
etc., e) maintaining or exerting personal control 
(questions, decisions, choices, etc.), f) manipulating 
outcomes, g) resolving issues, h) sharing fee lings 
and thoughts, i) building relationships [4]. 

When working with individuals who do not 
independently generate or program novel messages, 
the assistive programmer may begin by sensitively 
introducing the subject of private messages and 
options for maintaining confidentiality. All 
agreements related to the progr ammin g of such 
messages should be documented, reviewed for final 
^proval and kept on file. Potential topic categories 
for messages may then be presented auditorially or 
via a topic board. Some frequent categories for 
private messages include; legal matters, finanr.ial 
concerns, romantic partnerships, sexual matters, 
family issues, personal/emotional challenges, 
guardianship, abuse/neglect, discrimination, 
personal experiences, counseling dialogue, medical 
questions, and concerns regarding professionals and 
services. Before discussing the preferred content of 
messages within categories the individual who uses 
AAC should be provided the option of working ? 



with others to complete the programming process. 
It is possible that different preferred programmers 
will be identified for each category. In such cases, 
the primaiy programmer should attempt to instruct 
those people identified. Once messages and 
programmers have been indicated, strategies for 
incorporating the messages into AAC systems may 
be selected. Some potential options include: 

1. Masking- For messages which must be 
explicitly displayed on static co mmuni cation 
boards or device overlays, a simple covering of 
laminated, opaque paper may be used to mask the 
section of the board containing the messages. 
Velcro or tape tabs, metal or spiral rings, etc. may 
be used to secure the mask. If at all possible 
construction should allow for independent 
manipulation by the user. Instructions for use 
and/or a privacy warning may be placed on the 
superior surface of the mask. 

2. Dedicated Boards/Overlays/Levels- One or 
more communication boards, overlays, or levels of 
high technology voice output memory may be 
dedicated for the purposes of private messages. If 
multiple overlays/boards are used by the individual 
but assistance is required in manipulating them, 
private message displays should be stored in a 
container separate from other displays and should 
be labeled with instructions for use, including who 
the user gives permission to manipulate them. 

3. Alternative Coding- Individuals who use a 
particular encoding strategy (i.e. letter, icon, etc.) 
may find it preferable to use an alternative 
encoding strategy for private messages. For 
example, an individual who communicates general 
messages via letter or icon codes may use number 
codes for private messages only. The reference 
for these codes may be stored in a confidential file 
or committed to memory so that only the user and 
programmer are aware that they exist. Clinical 
experience with this method suggests that user 
recall for private message codes is particularly 
good. 

4. Password Approach- This is one basic strategy 
which can be utilized to implement private 
messages on computer-based communication 
devices. In this ^proach the system operator 
needs to enter a specific sequence of keys, letters, 
or symbols, without prompting, in order to gain 
access to a set of private messages. If this is an 
intended feature of a given system or device, of 
course, it could be made available as part of the 



RESNA*94 • Jane 17-22, 1994 



no 



PRIVATE MESSAGES IN AAC 



er|c 



firmware or operating system of the device. 
Otherwise, if a given device is sufficiently 
programmable or has adequate macro capabilities, 
it should be possible to devise a technique to 
implement this strategy for private messages. It 
can certainly be adtpted to different vocabulary 
organization techniques. In a dynamic screen based 
device, it could be used to restrict access to hidden 
pages or levels of messages. In systems based on 
traditional orthography (e.g. abbreviation 
expansion), it would probably require specific 
firmware or operating system support to designate 
certain sets of encoded words and sentences as 
inaccessible until the password has been entered, 
since it would not be feasible to restrict the use of 
any of the 26 letters available as encoding elements. 
In icon-based encoding systems (e.g. semantic 
compaction), the larger number of encoding 
elements makes it possible to restrict access to one 
or more icons and use these in sequences for 
private messages. 

5. Blank Key Approach- This approach, simpler 
but perhi^s not as secure as the password 
approach, is based on simply hiding or not labeling 
one or more of the encoding elements or page 
selections needed to access private messages. This 
solution is simple enough that it should be possible 
to implement on just about any system, and places 
minimal demands on the system operator for 
additional key activation or remembering abstract 
codes. By including a blank, private message, the 
likelihood that some other unfamiliar individual 
would find any of these messages is greatly 
reduced. If the device being used provides support 
to prompt the system operator in entering valid 
code sequences (e.g. icon prediction), it may be 
desirable to circumvent or disable this feature for 
the hidden coding elements. However, the 
selection technique may depend on the use of this 
feature (predictive scanning) in which case it would 
need to be re-enabled to access the private 
messages, somewhat reducing the efficacy of this 
rqrproach. 

Discussion 

The importance of including private messages in 
AAC systems should not be overlooked or 
minimized. Selecting, programming and accessing 
such messages presents a number of ethical and 
strategic challenges for people who rely on AAC, 
their assistive prograiraners and significant others. 
Programmers must present people who use AAC 
with comfortable opportunities to construct private 

RESNA ’94 



messages, respect their wishes and guidelines in 
the development of private message strategies and 
procedures, secure confidentiality, and avoid the 
inclination to edit or otherwise unduly influence the 
development of such messages. 

The authors have provided some basic, clinically 
tested guidelines for developing private messages 
and a set of suggested strategies for programming 
and concealing such messages in a variety of AAC 
systems. Further explanation and demonstration of 
these methods will be given in the presentation of 
the ptq)er. Additional research and development, 
cliniod testing, and consumer input are neeaed to 
increase user independence in creating and flexibly 
manipulating messages of a private nature. 

References 

[1] Bantgee, M., Sands, M., and Schwery, L. 
(1989) Affect and it’s communication in the closed 
head injury population while using augmentative 
communication devices. Proceedings of the 10th 
Southeast Augmentative Communication 
Conference, pp 7-12, Birmingham, Alabama. 

[2] Beukelman, D. and Mirenda, P. (1992). 
^^ifmientative and Alternative Communication: 
M^napement of Sever n Communication Disorders 
ip ChilHren and Adults. Baltimore, MD: Brookes 
Publishing Co. 

[3] Light, J. (1988). Interaction involving 
individuals using AAC systems: State of the art 
and future directions. Augmentative and 
Alternative Cnmmiinication .4. 66-82. 

[4] Sweeney, L. A. (1989). Engineering success 
with augmentative communication strategies. Short 
Course, Michigan Speech-Language-Hearing 
Association Convention, Traverse City, MI. 

Lynn A. Sweeney, Director, Alternative 
Communication Technology Center, Central 
Michigan University, Mt. Pleasant, MI 48859 

Clifford A. Kusbler, Technical Director of R&D, 
Prentke Romich Co., 1022 Heyl Rd., Wooster, 
OH 44691 




• Jane 17-22, 1994 



111 



DESIGN STRATEGIES FOR AAC SOFTWARE, 
DO THESE STRATEGIES COMPLY 



APPLIED TO THE BLISSPROCESSOR. 
WITH THE USERS NEEDS? 



Ir A.D.Hekstra. 

HANDICOM, Harderwijk, The Netherlands. 



ABSTRACT 

Communication has many appearances and is 
far from a static process. Disabled users who 
need tools for communication are putting a 
wide range of demands, not only of operatio- 
nal nature. Flexible designing strategies for 
the development of appropriate devices for 
AAC users are of crucial importance. In this 
paper a modular design approach and its 
application are verified against the primary 
goal: to meet user communication needs by 
AAC software. 



BACKGROUND 

In cooperation with the Bliss Communication 
Centre in Holland we have formulated design 
strategies for AAC software, focused on 
Bliss. These design strategies have been 
presented on the 3rd ICCHP conference, 
Vienna July 1992. Meanwhile, software has 
been developed for word processing Bliss 
documents as well as for speech generation, 
and different other Blissymbol related 
activities. That same spring, about 30 speech 
therapist conducted an intensive pilot study 
for about three months, using a prereleased 
version of the BlissProcessor. They sharpened 
and reformulated both user specific require- 
ments as well as therapist related desires. 

The software package has been completed in 
October 1992 and is now available in Dutch, 
English, German, and French languages; 
others are under preparation. 

Since the introduction, dozens of bliss users 
and therapists are working with the BlissPro- 
cessor in many different ways, emphasising 
the changing demands for communication. 



STATEMENT OF THE PROBLEM 

In general, a communication tool can be 
described as an intermediate between a user 
and his or her surrounding world. 



Bliss 




(technical) 




outside 


user 


< — > 


aid 


< — > 


'world' 






vocabulary 







Figure 1. The basic communication tool. 



For many reasons, it is preferable that 
communication aids are universal applicable. 
Developers and suppliers prefer fewer and 
more versatile devices above numerous very 
specific aids. Therapist and rehabilitation 
centres hesitate to invest money and/or time 
in new devices and software if it can be used 
only occasionally. 

But the group of AAC users who need such 
aids for communication is far from homoge- 
neous. They differin motor abilities, vocabu- 
lary, cognitive level, visual abilities and last 
but not least, in social environment. As we 
have seen in the past years, the therapeutic 
and parental support plays a major roll in 
acceptance and optimisation. 

The vocabulary of the user will likely grow 
or change gradually while the AAC user is 
maturing. In many cases, he or she will get 
some help to extend the symbol chart. While 
a communication device becomes personal 
and indispensable, it is unacceptable having 
the system to be returned to the manufacturer 
for a vocabulary upgrade. Therefor, a system 
should be flexible and easy to use. 



DESIGN 

In an early stage of the design, adaptation 
(how to operate and understand the tool) 
should be separated from the application 
(what the tool is used for). This separation 
has several advantages. First, it makes it easy 
to optimize an application (i.e. speech 
generation) to user specific needs, just by 
selecting a different input/display module 
and appropriate settings. Secondly, a modular 
approach can make the device multipurpose. 
Thirdly, future extensions can be added 
easily. 



RESNA *94 17-22, 1994 



112 



DESIGN STRATEGIES FOR AAC SOFTWARE, DO THEY MEET USER NEEDS? 



O 

ERIC 



The decoupling of adaptation from application 
has another advantage. As it brings homoge- 
neity towards the user, different kind of 
applications (training, speech or document 
processing) become easier to understand and 
operated. 

While application developers don’t have to 
focus on time consuming user specific 
adaptations, they will be motivated to create 
more and new facilities for communication. 
If standard hardware and software is used 
wherever possible, it offers a high degree of 
future compatibility and will minimise costs. 

It is clear that AAC software should cover 
several input strategies and devices, support 
different vocabularies, symbol sizing and 
background colouring to serve as a tool for 
communication, whether by speech, by paper 
or anything else. Such software should be 
modular as the need for adaptation and exten- 
sion will never end, due to the nature of what 
we call "a living language". 



adaptation aid application 




DIS = Display, 

INP = Input module, 
LIBR » Symbol library 

An example could be: 

A1 = Voice synthesizer 
A2 = Text-processor 
A3 » Training course 
A4 = Environment control 



A1..A4 = Applications 

II .. 15 = Interfaces 

SI. .S3 = Support modules 



51 = Bliss <--> Dutch 

52 = Blissymbolic printing 

53 = English <--> Bliss 



Figure 2. A modular concept for AAC software. 



any other application. New features may be 
added easily as the developer of an applica- 
tion does not have to bother how the user 
interacts with its device. He only has to 
comply with the specifications of the interface 
(15 in figure 2). 

DEVELOPMENT 

Input. All modules of the BlissProcessor can 
be operated by scanning, by pointing or by 
keyboard entry. Several scanning principles 
are supported, operated by one or two 
switches. 

Pointing can be done by hardware devices 
such as mouse, joystick, multiswitch etc., all 
capable for four directions and one or two 
buttons indicating "Yes" or "No". Keyboard 
users may profit from many shortcuts and 
additional search functions. 

Display. Apart form entry hardware, each 
user has his own vocabulary, symbol sizing, 
background colouring, command complexity 
level, selection strategy etc. A separate 
therapist/parent operated module enables 
modification of the vocabulary (chart) and all 
related settings. 

Applications. Several program modules have 
been developed, both for the user and thera- 
pist. These are: 

* BlissText, a word processor for Bliss- 
symbolic documents; 

* BlissTalk, a speech generator optimised 
for speech reproduction and synthesis; 

* BlissEditor, a module for the therapist to 
create new symbols fast and easily, and 
update them into the library; 

* BlissChart, to modify user charts and user 
specific settings. 

All program modules use the same symbol 
library and may invoke the same conversion 
routines to perform a communication task, 
such as printing facilities for blissymbols. 
Towards the user it appears that each module 
is completely designed for his or her personal 
demands: operation, vocabulary and feedback 
by display. 



In figure 2 the modular approach becomes 
visible. At the left side, the user interacts 
with its communication device by an optimal 
input/display interface. At the right, inter- 
action with the outside world is fulfilled by 
speech, written texts, environment control or 



System support. A PC, running MS-DOS is 
chosen. A reasonable priced widely available 
system. As the world market asks for smaller, 
portable robust computers, the AAC part of 
the society will only profit from it. 

127 



RESNA ’94 ^ June 17-22, 1994 



113 



DESIGN STRATEGIES FOR AAC SOFTWARE, DO THEY MEET USER NEEDS? 



EVALUATION 

In the beginning, therapist where a bit 
anxious to work with the BlissProcessor. 

As many options were available and quite 
some speech therapist had no computer 
experience at all, they wondered how such 
AAC-software could be useful. But soon, 
they discovered several advantages of the 
computer-aided training. For example: 

* A Bliss message becomes less volatile if 
presented the screen or on paper, offering 
facilities for training grammar. 

' Therapist and parents can divide their 
attention to more than one Bliss user in 
the room without loss of communication. 

Bliss users had less problems in acceptation, 
for them it was often the first time of their 
life they could make a letter on their own and 
send it to grandma dr tell a story without the 
constant need of someone standing besides the 
wheelchair. Once the input device was opti- 
mized (critical important!) and once they 
were able to choose symbols from their her 
own vocabulary, many users tend to become 
addicted to the computer. 

Both for teaching purposes and for writing 
purposes, the BlissProcessor is used inten- 
sively. Training material, poetry booklets are 
made and letters are mailed around. 

By the time this article was written, speech 
generation was of second importance, due to 
the fact that only very few users had a 
portable computer. Speech synthesis was not 
fully implemented into the BlissProcessor 
until early 1994. 



DISCUSSION 

As many users of the BlissProcessor are using 
different computers at different places, they 
face some problems. Although they can easily 
take along their own bliss* chart on diskette, 
user specific symbols are not available 
everywhere. 

Another problem arised as different 
therapists were modifying the symbol library 
at different computers, making it difficult to 
keep track on the home made symbols. 

Apart from these initial stragglings, the 
modular design strategies have worked very 
well for the BlissProcessor. Therapists enjoy 

. Xc Vi 4 

RESNA’94 • 



the possibilities of easy and fast making of 
new symbols, and the easiness of modifying 
the user’s symbol charts and the printing 
facilities. From the users point of view, the 
document processor offers a complete new 
way for communication they had never 
dreamed of. A rugged portable with speech 
support is the next step to look for. 

The flexibility of the multi-paged symbol 
charts offer the user access to an unlimited 
extendable vocabulary, independent of his 
physical limitations. Several users who have 
worked with about 400 symbols for many 
years are now demanding for more symbols. 
Until now, their physical limitation was 
limiting their vocabulary. 

As the modular design has proposed, all 
people involved in the blissymbolic 
communication can get access to more or less 
specific parts of the AAC device. The system 
is easily adaptable and it can therefor be 
matched to user specific needs. It offers 
support for different communication tasks 
opening pathways for future extensions. 



REFERENCES 

[1] Balkom, H.van, Welle Donker- 
Gimbrere, M, Kiezen voor Communi- 
catie (INTRO, Nijkerk 1988). 

[2] Bliss, Ch.K., Semantography (Blissym- 
bolics publications, Sydney 1965). 

[3] Hart, J.’t, Ontwerp en implementatie van 
een BlissInterface op de PC. (Twente 
University of Technology. 1991). 

[4] Hehner,B. et al. Blissymbolics for Use. 
(Din Mills, Ontario 1986). 

[5] Schaerlakens,A. et.al. leder zijncommu- 
nicatie. (ACCO,Leuven/Amersfoort 
1986) 

[6] Ir A.D. Hekstra; HandiBliss, a multi 
modular approach in program develop- 
ment. (ICCHP3 proceedings, 1992) 

[7] Drs J.L.M. v.d. Broek; Enquete 
Blissverwerker. (Kompagne, 1992) 

Doeko Hekstra, 

HANDICOM, 

Oranjelaan 29, 

3843 A A HARDERWIJK, 

The Netherlands, 

Tel: +31 3410 12629 

Fax: +31 3410 30602 

June 17-22, 1994 



114 



TRANSLATION PROBLEMS OF TERSE MESSAGES IN 
VOICE OUTPUT COMMUNICATION AIDS 



Jef&ey C. Mich^ and Bruce R. Bak^ 
Semantic Compaction Systems, Pittsburgh, PA 



ABSTRACT 

Computo*>based augmentative conununication 
(AQ systems use terse messages to communicate 
functic^ aspects. Translations of such terse 
messages from one natural language to anodic 
produce problems. Systematic effcHts used to 
produce terse prompts in one language have to be 
adapted when to anoth^ language. 

Felicitous results depend upon a thorough 
knowledge of the computer program and a 
knowledgeable native speaker of the target 
language. 



BACKGROUND 

Terse messages abound in life. They are found in 
public signs (STOP, YIELD, EXIT, etc.), posonal 
memos (call doc for 3ippL ), and computer program 
instructions (CUT, COPY, PASTE, etc.), among 
other places. These messages serve to reduce the 
amount of text of a longer message for space and 
time considerations. In one's native language, a 
reader of a terse message does not usually have 
many problems in deriving the original message. 
This paper wiU f^esent some of the problems 
which have arisen in translating the operatkMial 
systems of computers with terse messages into 
French, Spanish, Italian, German, and Swedish. 
Also, it will provide recommendations to others 
who are faced with the task of rendering computer 
systems comprehensible to speakers of other 
l^guages. Examples are provided from recent 
work done in translating the Prentke Romich 
Company Touch Talker™ (TT)/Light Talker™ 
vLT) operational systems. 

STATEMENT OF THE PROBLEM 

The messages relevant for the current discussion 
can be divided into separate groups: prompts 
(appearing in the display) and labels (on keys and 
s^Tpearing in the display). Labels in turn can be 
(hvided into functional and iconic categories. The 
prompts are limited by the numb^ of charact^ 
avail^le in the liquid crystal display. Both TT 
and LT displays allow 36 characters for prompt 
messages. The functional labels are hniited three 
lines of eight characters by the amount of space on 
each key. The icon labels are allowed eight 
characters by the software. As with most oth^ 



compute systems, the need for terseness is evident 
on the TT and LT. 

Problems arise wh^ such brief messages are 
translated from (Hie language to aiK>thCT. Two 
characteristic problems are ccnnmon. One is the 
creation of a new message in the target language 
which is substantially long^ than the original 
message and does not fit into, the allotted space. 

An example is END TO SAVE. The meaning of 
this statement is "select the END key to save a 
vcxabulary item." The initial Fr^ch translation 
became " APPUYER SUR FIN POUR 
SAUVEGARDER," a differmce of 21 characters! 

The seamd typical (Hobl^ is the mistranslation 
of a message owing to the translatcH‘'s 
misundo^tanding of the (Higinal terse message. 
The basic cause of such misunderstandings is a 
lack of knowledge of the operation system on the 
part of the translaUH*. A rath^ cryptic message 
from the TT/LT operation system, "FOR 
LOCKED MESSAGES; OK?" This function 
allows one to review messages whose icon 
sequences are in conflict with oth^ icon 
sequences. The translator originally rendoed the 
functicMi as "EMreCHER L'AOCES AUX 
MESSAGES; OK?" This literally means "prevent 
message access." This translation actually 
reverses the meaning of the frmcticHi. The problem 
here was the translatcv's lack of understanding of 
how the message frt into the system. 

Thus three processes take place. The first can be 
viewed as expanding the terse message to a more 
expository version of itself. The second step is 
translating the expanded message. The third step 
is rendering the expanded, translated message terse 
in the target language. 

The foregoing issues must be understocxl in the 
context of basic syntactic and semantic differences 
whk:h exist between even closely related 
languages. 

English has a special aptitude for terse messages 
because of its ability to use noun compounds. An 
example of this is the Italian name for the software 
called in English Words Strategy™. The Romance 
languages - Italian, Fr^ch, Spanish, Portuguese, 
Romanian, etc. do not permit nouns such as 
"w(Hxl" to be used descriptively. One cannot say 



.129 

RESNA ’94 • June 17-22, 1994 



115 



TRANSLATION PROBLEMS 



"Parole Strategia" in Italian. A correct rendering 
would have to be "La Strategia delle Parole." 

Unfortunately, even this rendering does not mean 
the same as Words Strategy^ means in English. 
The separatxMi of "Words" into a prepositional 
phrase substantially distrxts the notion conveyed in 
English. The meaning in Italian is quite close to a 
direct translation of the Italian phrase back into 
English: The strategy of the words. 

APPROACH 

Although significant diffoences exist annong 
languages, systematic emms used to produce terse 
messes in one language can be adapted when 
applied to other languages. Sevoal strat^es 
based on linguistics and common sense have 
proven useful. 

• Choose the shortest possible message 
A French version of Minqreak™ Words 
Strategy™ software uses an icon similar to an icon 
named CLIMB in the English vision which 
depicts a pair of snow covered mountain trqrs. 

The French icon adds a minivan in Croat of the 
riKMmtains. The two possible labd names were 
MOOTAGNE = "mountain" or CAMIONNETTE 
= "mini-van." The choice was obvious, 
MONTAGNE, not only because it was shrater 
than CAMIONNETTE but also because it only 
contains eight letters, the maximum allowed by the 
software. This pomits the storage of the whole 
label rather dian an abbreviation. 

• Appeal to established abbreviations 
The burden on the user of learning a new code is 
eliminated when conventional abbreviaticHis are 
used. Ablveviations like err/, cqps, rob, esc can 
sometimes be used across languages or their 
equivalents can be sought in the target language. 
For example Ctrl is understood in FreiKh, Italian, 
and Spanish. 

• Use acronyms Acronyms may be more 
widely used in some non-English languages than 
they are in American English. Acronyms are 
particularly abundant in French, Italian, and 
Goman among others. Acronyms are especially 
useful for long titles EMOS = Enhanced 
Minspeak™ Operating System. 

• Use telegraphic language The use of 
telegraphic language is universal on public signs. 

It is acceptable in computo prompts and labels 
even when grammatical rules of a given target 
language require the use of the {nepositions, 
articles, conjunctions, etc. Telegrtqthic language 
is often the most natural solution fm shortening - 
long translations. Some examples with their 



French translations are the following: ESCAPE 
STORAGE = SORTIR MODE MEMOIRE sordr 
de la mode m^moire; VOCABULARY 
UNREADABLE (the vocabulary is unreadaUe) = 
VOCABULAIRE HJLISIBLE; THEME HAS 
BEEN CLEARED = THEME EFFACE le thime a 
it£ ^aci. 

• Truncation This is an intuitive |MXx:ess 
which involves chopping off the end of a wrad, 
while leaving just enough for recognition. 
Truncation is effective in a range of languages. It 
must be noted, however, that syllaUe divisions are 
different even in closely related languages. All of 
the translations mentio^ in this paper employed 
saliency-based, wrxxl shortening extensively. 

Some French examples are: SECOND. 
{secondedre) = altouate overlay, CARACT. 
{caractire) = character, MAJUSC. {majuscule) = 
shift. 

• Si^ix reduction Though this process is 
similar to truiKatkm, it tqipeals qrecifically to the 
ability of the operator to recognize the (Higinal 
suffix in an abbreviated form. Suffix reduction 
should be ermsistent throughout the system to 
lessen the burden on the user of learning to 
understand this strategy. Some examines are: 
SELECT N {silictioti), FONCTN {fonction). 

• Vowel reduction Vowels can be 
omitted from wrmls, especially when they are 
found in unstressed positicHi, without creating 
recognition jnoblems. OVRLY= ovolay is an 
example from English which exploits this strategy. 
Vowels play different redes in die various 
languages. Though most English vowels are 
reduced to schwas in actual pronunciation, this is 
not the case with vowels in Romance languages. 
Because of different linguistic rules ftw vowels and 
because of the intuitive nature of recognition 
strategies, caution needs to be exercised in 
employing this technique even with the 
collaboration of a knowledgeable native speaker. 

An examfde of sevoal strategies employed at once 
can be seen in the translation for the English 
prompt ONLY WORKS IN CUSTOM 
OVERLAY 128 MODE. The French translation 
of this prompt became "fonctionne seulement en 
mode 128 acetate sur mesure..” Although this 
message is reladvely terse and difficult to interpret 
without a thorough knowledge of this system, it is 
still too long. The final {Hompt became: 

FCTNNE. SEUL. MODE 128 ACET. S.M. Suffix 
reduction, vowel omission, truncadcHi, and 
telegraphic language were all employed to inoduce 
a prompt requiring 32 of the 36 characters 
available on the second line of the TT/LT liquid 
crystal display. 



116 



RESNA ’94 



m 



une 17-22, 1994 



TRANSLATION PROBLEMS 



DISCUSSION 

The apiHoaches described were used extensively 
for translating functional and linguistic aspects for 
three diffment voice output communication aids 
(VOCA) into a series of European languages. It 
was found that principles used to inoduce 
terseness and comprehensibUity in English could 
be adapted to the various targeted Indo-European 
languages. The extension of these principles to 
Hebrew, Arabic, Chinese, and Japanese is planned 
as a part of future developments. 

The translation process is a team effort The need 
for at least one computo’-lit^ate native speaker of 
the target language on the team should not be 
underestimated. Translators must be fully trained 
in the system to insure that all messages are 
understood and accurately translated. ProUems 
arise from translators* lack of deep understanding 
of the opoation of the computer. Further, the 
native speaker needs to have some knowledge of 
the general purposes of a VOCA in order to 
perceive whether a shortened message "works." 
Prompts used for data bases, word processing, etc. 
can be very diffment from those us^ for 
spontaneous, interactive language generation. 

Perhaps as a last note, it may be best to regard 
translation not as translation but as recreating an 
operating system in another language. This will 
help to steCT the team away from "rote" word for 
word translation which can lead to endless human 
factor problems. 



Jeffrey C. Micher 
1000 Killamey Drive 
Pittsburgh, PA 15234 




121 



RESNA’94 • June 17-22, 1994 



117 



QUANTITATIVE INDICATORS OF COGNITIVE LOAD DURING USE OF 
A WORD PREDICTION SYSTEM 

Heidi Horstmann Koester and Simon P. Levine 

Rehabilitation Engineering Program and Graduate Bioengineering Program 

University of Michigan 



Abstract 

This study investigates the time cost associated with 
the cognitive processes performed during use of a 
word prediction system. Fourteen able-bodied and 
spinal cord injured subjects transcribed text with and 
without word prediction for seven test sessions. The 
extra time required to make word prediction selec- 
tions, as well as times for keypress and list search ac- 
tions, were measured as indicators of cognitive load. 
All subjects had slower keypress times during word 
prediction use, compared to letters-only typing, and 
spinal cord injured subjects had slower list search 
times than able-bodied subjects. 

Background 

The importance of cognitive load in determining user 
performance with word prediction systems is well- 
known. Task analyses have identified some of the 
component processes that may contribute to cognitive 
load, such as deciding when to search the prediction 
list and the visual search itself [1,2]. An important 
question continues to be how to quantify the effect 
and magnitude of this load. Measures of overall text 
generation rate provide a broad indication of the 
extent to which the cost in increased cognitive load 
countoacts the benefit of keystroke savings [3]. A 
more specific measure of cognitive load is extra 
selection time, defined as the additional time required 
to make each selection in a word prediction system 
relative to letter-by-letter typing. Extra selection 
time is assumed to be a direct reflection of the time 
spent on the additional cognitive processes. 

Current knowledge about extra selection time is 
limited. A theoretical analysis has estimated extra 
selection time at 1.22 seconds, using results from 
information processing psychology [2]. The one 
reported empirical measurement is consistent with 
this, showing a range of 0.9 to 1.5 seconds, but 
measurements were made only for able-bodied 
individuals under a limited set of conditions [3]. 

Research Questions 

This paper focuses on one aspect of an experiment in 
which the effects of a word prediction feature were 
studied with respect to over^l text gen^ion rate 
and cognitive load. While these two aspects are 
closely related, their differences warrant separate 
treatment Specifically the goal of this study is to 
provide further understanding of the extra selection 
time associated with word prediction use, by 
measuring it under a broader set of usage conditions 
than have been studied previously. Additionally, we 
hope to measure cognitive load more precisely by 
separating the time for the primarily motor act of 



I»essing a key fiom the i»imarily cognitive- 
perceptual act of searching the list. This is important 
in understanding the relative contribution of motor 
and cognitive factors to overall performance as well 
as a source of model parameters for future 
simulations of user performance. 

Methods 

Subjects. Fourteen subjects participated. All shared 
the following characteristics: at least some college- 
level education; high familiarity with the standard 
keyboard; no significant prior experience with word 
pr^ction; and no cognitive, perceptual, or linguistic 
impairments. Eight of the subjects were able-b^ed, 
while the remaining six had spinal cord injuries at 
levels ranging from C4 - C6. 

Interfaces . The "Letters-only” system involved 
letter-by-letter spelling on a standard computer key- 
board, and the "LettCTS+WP" used single letter entry 
augmented by a word prediction feature. A six-word 
prediction list with a fixed word order was used, pre- 
sented vocally in the top left comer of the screen. 
Able-bodied subjects used mouthstick typing, while 
subjects with spinal cord injuries used their usual 
method of keyboard access, which was mouthstick 
typing for two of the subjects and hand splint typing 
for the other four. 

Experimental Design . An alternating treatments de- 
sign was employed, in which subject performance 
with and without word prediction was recorded in 
each of seven test sessions. The keystroke savings 
provided by word prediction was fixed across Ses- 
sions 1-4 and varied in Sessions 5, 6, and 7. Three 
spinal cord injured and four able-bodied subjects 
were assigned to use one of two word prediction 
strategies to form four subject groups: SCIl, SCI2, 
ABl, and AB2. 

Training . In the first part of training, subjects were 
instructed in the text transcription task, and then 
practiced using the Letters-only system for six blocks 
of text (four sentences each). The second part of 
training introduced subjects to the Letters+WP sys- 
tem and their assigned strategy for its use. The rule 
for Strategy 1 was to search the list before every se- 
lection. The rule for Strategy 2 was to choose the 
first two letters of a word without searching the list, 
then search the list before each subsequent selection. 
For both strategies, a search was not required when 
the list was empty. All subjects practiced using their 
strategy for four blocks of text (4 sentences each), 
which was sufficient for each to use the strategy cor- 
rectly without prompting. 

132 



RESNA’94 • June 17-22, 1994 



118 



Cognitive Load and Word Prediction 



Testing. Each of the seven test sessions involved 
four sentences of warm-iq) using word prediction, an 
eight sentence test with word iwediction, then a two 
sentence typing test. Text blocks were drawn firom 
published typing tests [4] and revised to provide spe- 
ciflc levels of keystroke savings. Sentences were 
presented singly on index cards which remained visi- 
ble throughout transcription. Subjects had twenty 
seconds to read the sentence before an audio cue sig- 
nalled them to begin transcription. 

Data Collection. Subject behavior in each session 
was recorded on videotape. Additionally, all selected 
items were timed and stored by the software in real 
time. These raw data were filtered to remove events 
judged to be in any of the following three categories: 
text errors and error corrections; words not entered in 
a manner consistent with the assigned strategy; and 
"card reads", or times when the subject referred back 
to the text card during transcription, as identified 
through analysis of the videotape records. 



decreased significantly as subjects gained expaience 
with the Letters+WP system (p < 0.0005). Subjects 
who used Strategy 2 generally had lower extra selec- 
tion times, since fewer list searches were required, 
but the difference was not statistically significant. 




Dependent Measures . Extra selection time was 
measured in each session and defined as the 
difference in selection times between the Letters+WP 
and Letters-only systems for that session. Selection 
time was defined as the total time required for a test 
divided by the number of selections (i.e., keystrokes) 
made during the test 

To measure cognitive load more precisely, the overall 
act of making a selection with Letters+WP was 
partitioned into the motor act of pressing the key and 
the cognitive-perceptual act of searching the word 
list. Each selection in a test was labelled according to 
whether it involved a keypress preceded by a list 
search or a keypress with no list search [5,6]. For 
example, when using Strategy 2, the first two letters 
of every word involved no list searches, so they were 
labelled as a keypress only. The keypr^ time (t|j) 
was then measured by averaging the times for all 
keypress-only selections. The list search time was 
calculated by subtracting one t|^ from the time 
recorded for each list search-plus-keypress selection, 
then averaging the remaining times. 

Statistical Analyses . Statistical differences in the de- 
pendent measures were determined using a repeated 
measures ANOVA technique. The between-subjects 
factors were strategy and presence/absence of spinal 
cord injury, and the within-subjects factors were 
system and session. Statistical significance within 
each test was judged at a familywise p-value of 0.05. 






ERLC 

hfiiinniiffnrniaaa 



Results 

Extra Selection Time. .Figure 1 shows the extra se- 
lection times for the four subject groups. Spinal cord 
injured subjects had significantly larger extra 
selection times than able-bodied subjects, averaging 
0.910 seconds compared to 0.413 seconds (p < 
0.0005). For all subjects, extra selection time 



COPY AVAILABLE 



RESNA’94 • 



Fig. 1. Extra selection time with Letters+WP, relative 
to Letters-only. 

Kevpres.s-onlv Time . The average keypress times 
during use of Letters+WP are shown in Figure 2. 
Thae were no statistically significant differences 
between the groups, either on the basis of strategy or 
spinal cord injury. Session did have a significant 
effect (p=0.001), as keypress time improved an 
average of 17.7% from Session 1 to Session 7. 




Fig 2. Keypress times during use of Letters+WP. 

An unexpected result was that teyptess times during 
use of Letters+WP were significantly slower than 
during Letters-only typing (p < 0.0005), with an 
avoage diffraence of 23% (170 msec). The keypress 
slow-down was more pronounced for subjects with 
spinal cord injuries than those without, with a slow- 
down of 48% (270 msec) for spinal cord injured 
subjects, and 10.8% (94 msec) for able-bodied 
subjects. However, this difference was not quite 
statistically significant. 

133 

June 17-22, 1994 119 



Cognitive Load and Word Prediction 



List Search Time. Figure 3 shows the average list 
search times for the four subject groups. As with 
keypress time, strategy of use did not significantly af- 
fect list search time (p=0.058). Spinal cord injury, 
however, did have a significant effect (p < 0.0005); 
the list search times of subjects with SCI were an av- 
erage of 96.4% (560 msec) slower than the able- 
bodied subjects. For able-bodied subjects, list search 
time improved an average of 27.3% (180 msec) from 
Session 1 to Session 7 (significant at p < 0.(X)05). 
For spinal cord injured subjects, howevCT, list search 
time improved only 2.7% over these sessions, which 
was not significant (p=0.395). 




Session 

Fig 3. List search times during use of Letters+WP. 
Discussion 

These results suggest that using word prediction ex- 
acted a substantial cognitive cost for these subjects. 
Based on the extra selection time results, the execu- 
tion of cognitive and perceptual processes comprised 
between 30% and 60% of the time spent using the 
Letters+WP system. The sources of this cognitive 
load and possible reasons for the differences seen 
between subject groups are discussed below. 

List search time was a major contributor to cognitive 
load, measuring at least several hundred milliseconds 
for all subjects. However, it is unclear why the list 
search time for spinal cord injured subjects was so 
much larger than for able-bodied subjects. This may 
have been due to a difference in searching styles; 
perhaps the spinal cord injured subjects performed 
more thorough searches, while the able-bodied sub- 
jects used anticipation to skim the list on occasion. A 
second possibility stems from the fact that the tech- 
nique used to measure list search time also captured 
the time for other processes, such as ' deciding 
whediCT to search or verifying accuracy of a selec- 
tion, if and when they occurred. These processes 
contribute to the cognitive overhead involved in 
monitoring and guiding overall activity, and this 
overhead may have been greater for spinal cord 
injured subjects. This would not be that surprising, 
since they had much more prior experience undo' the 



Letters-only condition than the able-bodied subjects, 
which would increase the cognitive difficulty of 
switching to the Letters+WP system. 

Cognitive overhead may also account for the result 
that keypress time was slower with the Lettos+WP 
system. If keypress time had measured only the 
motor component of item selection, as intended, its 
duration should have been essentially the same with 
and without word prediction. That it was not 
suggests that cognitive overhead was present even 
during selectiois that did not involve a lik search [2]. 
The relatively large keyjHess slow-down seen for 
spinal cord injured subjects provides furdier evidence 
of greater cognitive overhead for these subjects. 

Certainly future woric is necessary to address these 
utuesolved issues and to examine cognitive load un- 
der different sets of conditions. This study provides a 
methodological starting point for future wcsk as well 
as intriguing initial results. Continued progress to- 
ward understanding cognitive processes and their as- 
sociated time costs is critical in order to clarify how 
cognitive load impacts ovoall user performance and, 
ultimately, to determine ways of reducing its effects. 

References 

1. Soede M, Foulds RA. (1986). Dilemma of 
prediction in communication ai^. Proc. of 9th 
RESNA Corf., Wash., D.C.: RESNA, 357-359. 

2. Hwstmann HM, Levine SP. (1991). The 
effectiveness of word prediction. Proc. of 14th 
RESNA Corf., Wash., D.C.: RESNA, 100-102. 

3. KoestCT HH, Levine SP. (in press). Learning and 
performance of able-bo^ed individuals using 
scaruiing systems with and without word 
prediction. Assistive Technology. 

4. Lessenbeify D. (1975). College Typewriting. 
Cinciimati: Southwest^ Pub. Co. 

5. Card S, Moran T, NeweU A. (1983). The 
Psychology of Human-Computer Interaction. 
Hillsdale, NJ: Erlbaum Associates. 

6. Olson JS, Nilsen E. (1988). Analysis of the 
cognition involved in spreadsheet software 
interaction. Human-Computer Interaction, 3:4, 
309-350. 

Acknowledgements 

Many thanks to Aram Sherman for his assistance in 
perfmming this study and to the subjects for their 
participation. This work was supported by the 
National Science Foundation, the University of 
Michigati Rackham School of Graduate Studies, and 
the U-M Rehabilitation Enginening Program. 

Address 

Hddi Koesto* ' 

1C335 University Hospital" 

Ann Arbor MI 48109-0032 
(313) 936-7170 
Internet: hhk@umich.edu 



RESNA ’94 • June 17-22, 1994 



120 



Predictive Letter Scanner for Augmentative Communication 



Karen Hamilton 

Applied Science and Engineering Laboratories 
Alfred I. duPont Institute/University of Delaware 



ABSTRACT 

Communication rate remains a primary concern for 
people using AAC systems. In particular, single- 
switched scanning approaches, while useful for people 
with severe physical impairments, can be extremely 
slow. Letter prediction has been previously suggested 
as a rate enhancement technique. However, this tech- 
nique is not widely available. This paper explores the 
possible reasons for this and suggests an alternative 
implementation that may promote letter prediction use 
with scanning devices. 

INTRODUCTION 

Currently, scanning devices are available that provide 
basic assistance to people with severe speech and 
physical limitations. Individuals who benefit from 
scanning based devices typically are unable to use oth- 
er selection techniques such as direct selection (4). Un- 
fortunately, the slow rate of communication 
achievable using a single switch scanning device can 
cause the user and listener a great deal of frustration. 
The necessity of improving the single switch commu- 
nication device becomes clear when comparing the 
communication rate of a non-disabled person to that of 
a person with a severe expressive communication dis- 
ability. In normal speech a non-disabled person can 
communicate at a rate of 200 words per minute and 
type at approximately 55 words per minute (2). In con- 
trast, individuals with disabilities typically achieve 
communication rates on the order of 2 to 10 words per 
minute (7). For alphabetic scanning systems, decreas- 
ing the average selection time per letter will effectively 
increase the rate of communication. 

The goal of this research is to increase the communi- 
cation rate achievable with a single switch scanning 
device while not requiring increased physical dexterity 
of the user. Letter prediction is a rate enhancement 
technique particularly useful in scanning-based sys- 
tems. The implementation as explained in this paper 
provides a method of decreasing both the average 
number of scans and average number of selections per 
letter while minimizing the cognitive and perceptual 
burdens for the user. 

BACKGROUND 

Development of a rate enhancing technique for use 
with a single switch row-column scanner was initiated 
by the Biomedical Engineering Center at Tufts Uni- 
versity (2). This research resulted in the development 
of the Tufts Interactive Communicator (TIC) and the 
Anticipatory TIC (ANTIC). The TIC ordered the dis- 



P 

Jun 



play according to letter frequency in the English lan- 
guage. ANTIC expanded this concept by considering 
language redundancy to predict the most probable six 
letters based on the previous two letters typed. The six 
positions at the top-left of the display were dynamic 
and held the predicted letters. The remainder of the 
display consisted of the entire alphabet arranged ac- 
cording to frequency of occurrence. 

Further exploration of predictive scanning schemes 
has been done at Northwestern University. The Micro- 
DEC n used the statistical frequencies of language to 
improve text entry efficiency (8). The method of letter 
prediction used in the MicroDEC II utilized bigram 
frequency statistics to order the letters on the dynamic 
display according to their probability of following the 
previous character. The Portable Anticipatory Com- 
munication Aid (PACA), also developed at North- 
western University, utilized a dynantic display 
reordering scheme based on bigram frequencies as 
well (9). 

While letter prediction offers theoretical rate enhance- 
ment as high as 40%, the technique has not enjoyed 
wide-spread use. One possible reason is that these sys- 
tems made predictions by reorganizing the scanning 
matrix. This poses an additional cognitive perceptual 
burden on the user. In fact, experimental testing of the 
ANTIC showed that subjects experienced a great deal 
of mental exhaustion after 30 minutes of use. The de- 
cline in performance was attributed to the use of dis- 
play reorganization to present predictions (5). In 
contrast, the system described in this paper imple- 
ments letter prediction in a way that keeps the letter ar- 
rangement static. 

IMPLEMENTATION 

Like its predecessors, the Predictive Letter Scanner 
takes advantage of the statistical redundancy of the En- 
glish language to determine predictions for the most 
probable character to follow the previous selection(s). 
The predictive scannw utilizes a static ordering of let- 
ters based on frequency of occurrence. Unlike the dy- 
nantic displays utilized in the ANTIC, MicroDEC n, 
and PACA, predicted items are scanned in their fixed 
positions. By keeping the vocabulary configuration 
fixed, some potential for user confusion is eliminated. 
Use of a static display also decreases the burden on the 
user of learning item positions on numerous scanner 
display configurations. Therefore, the visual search 
time as well as the number of missed letters is de- 
creased. Finally, Barker’s clinical observations sup- 
port the use of a static display (1). She observed that 
experienced scanners look at the desired letter and se- 



• June 17-22, 1994 



RESNA ’94 



121 



Predictive Letter Scanner 



lect it once it is highlighted. 

The Predictive Letter Scanner software is implement- 
ed using LASO (Library of Adaptable Software Ob- 
jects), a C++ based software development library 
developed at the Applied Science and Engineering 
Laboratories (3). The current implementation runs on 
MS-DOS compatible computers. The letter.prediction 
module is incorporated into a previously developed 
row-column scanner. The vocabulary set consists of 
the 26 alphabetic characters, space, punctuation, and 
control instructions. Single switch row-column scan- 
ning is done by highlighting each row in succession 
until the user selects a row with a switch closure. The 
highlighted marker then traverses the row until the 
user selects an item with a second switch closure. 
Prediction Generation 

A primary concern in developing an effective rate en- 
hancement module is obtaining reliable predictions. If 
the desired letter is not among the predictions, the user 
must wait while each letter predicted is linearly 
scanned prior to beginning a row-column scan. For 
this reason as well as considering memory constraints, 
it is important to carefully choose the optimal number 
of predictions to present. Figure 2 shows the prediction 
reliability for the first six presented items. 



Reliability of predicted tetters 




Figure 2: Frequency of correct prediction for 
the first six predicted letters. 

Unlike its predecessors (which used bigram and tri- 
Rrani frequencies), the Predictive Letter Scanner uses 
up to three previously selected letters to predict and 
present the most probable next letter.^ The Brown Cor- 
pus, a text consisting of over one-million words, was 
used to generate these frequencies. 

Basing predictions on up to three previously selected 
vocabulary items increases the percentage of letters 
correctly predicted. In simulation experiments of pre- 
diction effectiveness using the Carterette Corpus, a 
text consisting of 77,800 words, the first letter predict- 
ed was correct 44.8% of the time.^ Bigram prediction 



1. The PACA system supplemented letter prediction with 
word prediction after the first two letters of the word were se- 
lected. Consequently, quadgrams would probably not have 
improved their prediction performance. 

2. Predictions based on up to three characters entered in cur- 
rent word only. 

■ 

; f. 

RESNA’94 



was correct only 27.9% of the time. 

Basic Operation 

Incorporating the prediction module into the row-col- 
umn scanner involves instructing the basic scanner to 
communicate with the rate enhancement module. Each 
selection is reported to the prediction module as well 
as presented to the user via the screen. Predictions are 
indicated by highlighting each item in its fixed posi- 
tion. If none of these are the letter intended by the user* 
the system proceeds through a row-column scan. 

The system is capable of predicting spaces and punc- 
tuation as well as characters. This is very useful when 
considering common word endings such as ing. In 
most instances, the vocabulary item that the us;.r wants 
is a space. By considering this at the time of prediction 
generation the letter predictor will begin by presenting 
a space to the user. An additional feature is tihe predic- 
tion of a backspace. This occurs when less than three 
predictions are available to scan. The feature is includ- 
ed to alert the user of a potential misspelling or entry 
error and give immediate opportunity for correction. 
The system allows the user to define the row-column 
scan rate. This allows users to adjust scan rate as they 
become comfortable with the vocabulary configura- 
tion. Because prediction scanning is not as systematic 
as row-column scanning, the reaction time may be 
slightly greater. Therefore, the system allows users to 
set a different scan rate for predictions. 

During development a question was raised about the 
reliability of predicting the first letter of each word. 
The conclusion was to allow the user to decide wheth- 
er predictions based on no previous letters would be 
helpful in enhancing the rate of communication. Addi- 
tionally, the user may turn the prediction off. 
Alternative Method 

The Predictive Letter Scanner described above reduc- 
es scan steps and switch closures. An alternative meth- 
od investigated presented the predictions first as a 
selectable group. In this case, the number of switch 
closures was the same as in simple row-column scan- 
ning, but the number of scan steps was reduced. Such 
a system would be beneficial to a person who has more 
controlled ability to operate a switch. When using 
Grouped Prediction, if the desired letter was not pre- 
dicted the user could avoid unnecessary scanning by 
immediately beginning a row-column scan. 

RESULTS 

The Predictive Letter Scanner research studied the 
ability of letter prediction to provide an increase in the 
rate and ease of communication above that achievable 
with a single switch row-column scanner. Simulation 
experiments indicate that our goal of enhancing the 
rate of communication, while not requiring increased 
physical dexterity, was successfully reached. 

Simulation Results 

The Carterette Corpus was used in simulation experi- 
ments to determine the effectiveness of the prediction 

•June 17-22, 1994 



122 




Predictive Letter Scanner 



module. Predictive systems which reorganize the visu- 
al interface are concerned only with decreasing the 
scanning actions needed to reach the desired selection. 
By linearly presenting each predicted item, the Predic- 
tive Letter Scanner can decrease the number of switch 
closures as well as the number of scan steps necessary. 
Simulations compared the Predictive Letter Scanner 
with a single switch row-column scanner to determine 
its improvement in scan steps and switch closures. Ta- 
ble 1 presents the improvement results for prediction 
of all letters as well as linear prediction excluding the 
first letter. The results of grouped prediction indicate 
that a user might experience 40% rate enhancement. 

l^ble 1: Predictive Letter Scanner 
Improvements 





Scan Step 
Improvement 


Switch Closure 
Improvement 


Linear Prediction 
(including first letter) 


22.3% 


34.9% 


Linear Prediction 
(excluding first letter) 


30.2% 


31.3% 


Grouped Prediction 


40.1% 


0% 



RATE ENHANCEMENT METHODS 

A number of methods have been presented to enhance 
the communication rate of an AAC row-column scan- 
ning device. The effectiveness of these methods was 
measured by considering the letters per minute im- 
provement using the simulation results compiled from 
the Carterette Corpus. Calculations were done by 
holding the scan rate constant and varying the switch 
closure time. The results show a correlation between 
the Grouped Prediction rate enhancement technique 
and the time taken to make a selection. The Linear pre- 
diction methods are much less effected by the variation 
of selection time. In fact, as selection time increases 
the benefit of using a Linear prediction scan rises. 

The theoretical results indicate that if it were possible 
to determine the user switch closure time the method 
providing the best rate enhancement could be used. 
Results indicate that all methods produce significant 
savings in time and effort. The method that is best de- 
pends on the user’s switch closure time. Determining 
the switch closure time may prove to be a difficult 
task, therefore, it is expected that the most widely used 
method will be linear prediction without prediction of 
the first letter of each word. The improvement of this 
method remains relatively constant as the switch clo- 
sure time varies making it attractive both to proficient 
switch users as well as users who require a greater 
switch closure time. 

CONCLUSIONS 

The incorporation of a letter prediction technique into 
a single switch row-column scanner is effective in its 
goal of increasing the rate of communication while not. 

OPY AVAILABLE RESNA’94 • 



requiring additional motor capabilities of the user. Fu- 
ture enhancements of the system include speech output 
and continued investigation of additional selection 
techniques. In addition, experiments with subjects will 
be conducted to estimate the rate enhancement that can 
be expected in practical use. 

ACKNOWLEDGEMENTS 
Funding for this research was provided by a Rehabili- 
tation Engineering Center, Grant Number H133E 
80015 from the National Institute on Disability Re- 
search, U.S. Department of Education. Additional sup- 
port was provided by the Nemours Foundation. 
Special thanks to everyone involved with the Architec- 
ture Project. 

REFERENCES 

(1) Barker, M. R., Input Systems for Integrated AAC 

and Power Mobility Devices, Presented at the 
1992 Augmentative and Alternative 
Communication Conference. Philadelphia, PA. 

(2) Crochetiere, W. /., Foulds, R. A., Sterne, R. G., 

(1974). Computer-Aided Motor Communication, 
1974 Conference on Engineering Devices in 
Rehabilitation (1-5). Boston, MA. 

(3) Demasco, P., Ball, J. E., Ty vand, S., Blodgett, D., 

Bradley, W., Dunaway, J., Kazi, Z., (1992). 
Towards Mc^ular AAC Software: An Object- 
Oriented Architecture, Proceedings^of the 15th 
Annual Conference^oaRehabilitationTechnologv 
(1 19-121). Toronto, Canada. 

(4) Fishman, I., (1987). ElectrQnic_Communication 

Aids . College Hill Press, Boston, MA. 

(5) Foulds, R. A., personal communication. 

(6) Foulds, R. A., Baletsa, G., Crochetiere, W. J., 

Meyer, C., (1976). The Tufts Non- Vocal 
Communication Program, Proceedings of th(L 
1976 Conference on Systems and Devices foUhe 
Disabled (14-17). Boston, MA 

(7) Foulds, R. A., (1980). Communication Rates for 

Nonspeech Expression as a Function of Manual 
Tasks and Linguistic Constraints, Proceedings of 
the 3rd Intemedon^ Conference on 
Rehabilitation Engineering (83-87). Toronto, 
Canada. 

(8) Heckathome, C. W., Leibowitz, L., Strysik, J., 

(1983). MicroDEC II - Anticipatory Computer 
Input Ad, 6th Annual Conference on 
Rehabilitation Engineering (34-36). San Diego, 
CA. 

(9) Heckathome, C. W., Leibowitz, L. J., (1985). 

PACA: Portable Anticipatory Communication 
Ad. Proceedings of the 8th Annual Conference, 
on Rehabilitation Technology . (329-331). 
Memphis, TN. 

(10) Heckathome, C. W., Voda, J. A., Leibowitz, L. J., 
(1987). Design Rationale and Evaluation of the 
Portable Anticipatory Communication Ad - 
PACA, Augmentative and Alternative 
Communication , vol. 3, num. 4 (170-180). 

ADDRESS 
Karen Hamilton 
25617 GA Tech Station 
Atlanta, GA 30332 
email: hamilton@cc.gatech.edu 

1 37 

June 17-22, 1994 



123 



COMMUNICATION INTERACTION 

INVOLVING A YOUNG AUGMENTATIVE COMMUNICATION DEVICE USER 

AND HER PARTNERS 

Annette T. Stagge, Ed.D. 

University of Cincinnati 
Department of Special Education #2 
Cincinnati, OH 45221-0002 



ABSTRACT 

The purpose of this study was to describe the 
communication patterns of one augmentative 
communication device user and her family partners 
as they interacted with one another. 

Ethnographic techniques were employed to 
capture the verbal and nonverbal behaviors. Micro- 
analysis of the participants’ communication behav- 
iors during various events indicated that each 
partner demonstrated a variety of partner character- 
istics. The results of this study show that when 
communication behaviors are analyzed within 
natural contexts for an extended period of time, 
family partners demonstrate a variety of communi- 
cation attitudes, functions, styles, and intents. This 
model of Partner Characteristics can be used toi 
yield a more accurate picture of the augmentative' 
communication users communication abilities. 



BACKGROUND 

This study investigated the partner character- 
istics, conversational functions and styles, and 
communication intents of three partners interact- 
ing with an augmentative communication device 
user, and how these factors affected interaction. 
While the partner’s characteristics were the focus of 
the study, the augmentative communication user’s 
contributions and how they affected the partner 
interaction were included in the analysis. 

OBJECTIVE 

Since this study was motivated by the assump- 
tion that augmentative and alternative communi- 
cation users are active communicators and are able 
to convey a variety of communication functions 
and intents, the following research questions were 
fomulated to guide this study. 

1. ) What opportunities are provided for the 
augmentative communication user to interact with 
a variety of partners in a variety of settings? 

2. ) What partner characteristics are demonstrated 
by each of the participants in each of the interac- 
tions identified by this study? 

3. ) What constitutes the primary functions of g;ach 



of the interactions identified by this study? 

4.) What are the communication intents that make 
up the linguistic environment provided by each of 
the participants in each of the interactions ident- 
ified in this study? 

The primary subject of the study, Elizabeth, was 
6 years of age. She attended regular kindergarten 
and received physical, occupational and speech 
therapy within her educational setting. Her primary 
diagnosis was cerebral palsy with a severe expres- 
sive language delay. Elizabeth frequently vocalized, 
spoke in one-word utterances, gestured, signed, and 
used her Touch Talker while playing and interacting 
with her family partners. Mother and Father share 
in the responsibility of Elizabeth’s daily care, play, 
education, and programming of her Touch Talker. 
Elizabeth’s sister, Samantha, was 9 years of age. 
Samantha interacted daily with Elizabeth. 

APPROACH 

The data were collected over a seven month 
period through recording extensive field notes, 
videotaping naturally occuring interactions, and 
informal interviews with the participants (5), (7), 
(8). The data were collected in a variety of locations 
including the home, the school, a hair salon, the 
park, and the market. Only the data collected in the 
home with Elizabeth’s family members (Mom, 
Dad, and Samantha) were analyzed for this study. 
Three interactive episodes were chosen for in-depth 
study because they represented the various com- 
munication styles of each partner. A “Take Off’ 
game (variation of Backgammon) interaction 
between Elizabeth and her father was considered to 
be representative of the father’s communication 
styles of informing and directing. A bookreading 
episode with Elizabeth and her mother “ A Car Trip 
for Mole and Mouse” represented the mother’s com- 
munication styles of informing, initiating, and co- 
participating. A pretend play-grocery shopping 
episode with Elizabeth and her sister, Samantha, 
represented the sister’s communication styles of 
directing and modeling. In each episode, Elizabeth’s 
communication styles included co-participating 
and initiating, j 




124 



RESNA ’94 • June 17-22, 1994 



augmentative communication 

DISCUSSION 

The data clearly illustrated that Elizabeth par- 
ticipated in a variety of activities with familiar 
partners, in a variety of communication contexts. 
Analysis of the field notes and videotaped data 
indicated the presence of three partner character- 
istics: sensitivity, responsivity, and consistency (9). 
Each participant adjusted and adapted their com- 
municative behaviors to one another; thus providing 
a responsive communication environment across 
context. Nested in these partner characteristics 
were the partner’s conversational functions of 
didactic talking, communication anteing, and 
communication terminating. The communication 
interaction consisted of communication styles and 
functions which shaped the partner’s communica- 
tion intents, which made up the linguistic com- 
munication environment, used in the interaction. 

The overall results of this study contradict 
previous research. The literature review suggests 
that in those studies partners are controlling, 
dominant, and occupy the majority of the conver- 
sational space (2), (4). The past research also 
described augmentative communication users as 
being passive participants and not contributing to 
the social interaction (1), (3), (6). The results of this 
study showed that when communication behaviors 
are analyzed within natural contexts over an 
extended period of time, functions, styles, and 
intents. These results highlight that communication 
intents and styles are partner and context specific. 
Analysis in this study also revealed that Elizabeth 
was an active communicator during all activities 
with all partners. She also demonstrated the flex- 
ibility and adaptability necessary to actively par- 
ticipate in the communication episodes examined. 
In light of these research findings, this model of 
Partner Characteristics can be used to guide future 
research to see if more naturalistic studies will yield 
a more positive view of the communication abilities 
of augmentative communication users. 



REFERENCES 

1. Basil, C. (1992) Social interaction and learned 
helplessness in disabled children. Augmentative and 
Alternative Communication, 8^, (3), 188-189. 

2. Buzolich, M. & Wiemann, J. (1988) Turn taking in 
atypical conversation: The case of the speaker- 
augmented communication dyad. Journal of Speech 
and Hearing Disorders , 47 , (3), 281-287. 

3. Harris, D. (1982) Communicative interaction invol- 
ving nonvocal physically handicapped children. 



Journal of Speech and Hearing Disorders , 2, (2), 
21-37. 

4. Light, J., Collier, B., & Fames, P. (1985) Communi- 
cative interaction between young nonspeaking phys- 
ically disabled children and their primary car^vers: 
Part 1-Discouise pattans. Augmaitative and Altenative 
Communication , L (2), 74-83. 

5. Lincoln, Y. & Guba, E. (1985) Naturalistic inquiry . 
London: Sage Publications. 

6. Rhyner, P., Lehr, D., & Pudias, K. (1990) An 
analysis of teacher responsiveness to communicative 
intentions of preschool children with handicaps. 
T jngimge. Speech, and Hearing Services in Schools, 

(2), 91-97. 

7. Spr^ey, J. (1979) The ethnographic interview . 
P hilade lphia: Pa, Holt, Rinehart, and Winston, Inc., 
Publishers. 

8. Spradley, J., (1980) Participant observation. N.Y.: 
Holt, Rinehart, and Winston, Inc., Publishers. 

9. Wilcox, M. J. (1989) Partners: The other side of 
communication progr amming . The Clinical Connec- 
tion. Winter, 6-7. 



Annette Thieman Stagge, Ed.D. 
University of Cincinnati 
Department of Special Education #2 
339 Teachers Collie 
Cincinnati, Ohio 45221-0002 



139 



RESNA *94 • June 17-22, 1994 



125 



DEVELOPING AAC SYSTEMS THAT MODEL BSTELUGENT PARTNER INTERACTIONS: 

METHODOLOGICAL CONSIDERATIONS 

Peter B. Vanderheyden^ Christq)her A. Pennington^ Denise M. Peischl^ Wendy M. McKnitt^ 
Kathleen F. McCoy^ Patrick W. Demasco^ Hans van Balkom^, and Harry Kamphuis^ 
^Applied Science and Engineering Laboratories, University of Delaware/A.I. duPont Institute 
^IRV, Institute for Rehabilitation Research, Hoensbroek, Netherlands 



ABSTRACT 

Augmented conversations are interactive. In a pilot 
study, we analyzed the transcripts of students with 
cerebral palsy describing pictures to their therapists. 
Described herein are some of the pattmis observed 
during these interactions, and how they may reflect 
more general features in augmented conununication. 
We suggest that some of these patterns should be 
modelled within future intelligent AAC systems. 

BACKGROUND 

Some computar-based augmentative and alternative 
commumcation systems are able to utilize knowledge 
of syntax, semantics, and vocabulary in ordw to facili- 
tate the production of complete and correct sentences. 
To date, a number of studies have discussed aspects of 
the convosational nature of augmented interactions, 
but little attention has been paid to how tl.ey might be 
integrated into the AAC system itself. 

Conversation is often a cooperative, bi-directional, 
and multimodal process of constructing and exchang- 
ing information. In the context of AAC, a conversa- 
tional partna* often becomes actively involved in 
constructing the augmented speaker’s message (1). 
The partner may ask questions, repeat part of the aug- 
mented speaker’s utterance, or simply nod and smile 
in agreement This feedback may in turn affect the 
itiessage being produced by the augmented speaks*. 

Computer-based AAC systems currently “see” only 
the words the user selects. The less information the 
user provides as input, the less the likelihood of accu- 
rate ouqtuL Studies with manual AAC systems sug- 
gest however, that other modes of communication 
may be preferred. For example, children with cerebral 
palsy chose to use vocalizations, gestures, or eye gaze 
as modes of communication far more often than their 
manual symbol boards in interactions with their moth- 
ers or their speech thersqrists (2).These alternate 
modes of communication may te critical to fully 
understand augmented interactions. Already, the use 
of gestural recognition is being considered for future 
AAC systems (3). 

In order to develop truly intelligent AAC systems, we 
must understand and address such charactoistics of 



conversational interactions between an AAC usct and 
a partner. 

STATEMENT OF THE PROBLEM 

The goal is to design an augmentative communication 
system that provides common interactional fealures of 
a conversation between a person using a communica- 
tion aid and a conversational parmer. 

The first step has been taken in defitting criteria for 
such a system by analyzing the data firom a pilot study, 
and identifying a number of patterns that occurred 
during the dyadic interactions. 

APPROACH 

Pilot Study. The pilot data was collected by transcrib- 
ing videos originally recorded by van Balkom (4). 
Adolescent students with c^bral palsy described pic- 
tures in a children’s book to their primary speech ther- 
rqnsts, using their own manual symbol charts. Four 
such adolescent-therapist dyads were videotaped and 
analyzed. 

Each student was instructed to describe the pictures as 
if telling a story to youngs children. The therapist 
was instructed to repeat each word as it was selected 
by the student, paraphrase the sentence when it was 
completed, and then ask the student for confirmation 
that the paraphrased intopretation was correct. A sin- 
gle camera was used to videotiqie both the student and 
the therapist Students took between 11 minutes and 
one hour to retell their stories. 

Transcription system. In an effort to capture as much 
of the multimodal content of the interactions as possi- 
ble, the following vocalizations and gestures by both 
students and therapists were recorded: 

• vocal productions (both words and non-words) 

• hand and arm gestures 

- pointing at the commuitication board 

- pointing at the storybook 

- pointing elsewhere 

- unsuccessful attempt to turn the page 

- successful attempt to turn the page 
• miscellaneous gesture 

• facial expressions 



- smile 



^ '140 

RESNA’94 • June 17-22, 1994 



126 



Modelling Partner Interactions 

- miscellaneous facial expression 
• head gestures 

• head nod 

• head shake 

- looking elsewhae (not at board, storybook, or 
therapist) 

- miscellaneous head gesture 

Eye gaTft was not recorded, because experimental 
conditions did not allow accurate judgment of eye 
gaze direction. 

DISCUSSION 

Initial analysis of the data has indicated several inter- 
esting features and patterns of interaction that will be 
investigated further. For the sake of discussion, these 
observations are grouped into sevraal categories, none 
of which should be considered exhaustive. 

Co-construction. One intriguing behavior observed 
was that the thertqnst often repeated a sequence of the 
student’s selections before a sentence was completed. 
The form of this repetition can be described as a func- 
tion of two dimensions: degree of incrementalrty, and 
degree of int^retation. 

Table 1: iDcrementality vs. Interpretation 





Student 


Non- 

Incremental 


Incremental 




boy 


boy 


boy 


Non- 








Interpreta- 


girl 


girl 


boy girl 


tive 










walk 


walk 


boy girl 
walk 




boy 


a boy 


a boy 


Interpreta- 

tive 


girl 


a girl 


aboy anda 
girl 




walk 


walking 


aboy anda 
girl are 
walking 



To illustrate (Table 1), the student might select the 
symbols ‘boy’, ‘girl’, and ‘walk’. The therapist would 
echo “boy” after the first word, ^th a non-incremen- 
tal/non-interpretative strategy, the ther^ist might also 
simply echo “girl” after the second word. With an 
incremental/non-interprefative strategy, the ther^ist 
might say “boy girl”. With an incrementalAmterpreta- 
tive strategy, the ther^ist might say “boy and girl”, or 
“the boy and the girl”, or “they”. 



RESNA *94 



These strategies were not observed consistently: a sin- 
gle dyad might contain several strategies, or combina- 
tions of strategies. Interestingly, these strategies were 
observed despite the instructions that therapists should 
repeat the symbols only as they were selected, and 
interpret them only at the end of the sentence. This 
suggests that therapists found it natural to respond this 
way in the communicative situation. 

Word Find. When the student was unable to find a 
desired word on the communication board, or did not 
know a word, several strategies were used. The stu- 
dent might select a similar or related word, or point at 
the storybook, or use gestures to express the idea. For 
example, to express the word for sweeping, students 
moved their hands in a sweeping motion. Strategies 
employed by the ther^ist included guessing at the 
elusive word, examining and describing the picture in 
the storybook or asking the student to spell the word 
if it was known but did not appear on the communica- 
tion board. 

Conversationoi Repair. The student indicated that the 
therapist’s interpretation was incorrect in a variety of 
ways, including head shakes, uttering “no”, or point- 
ing at “no” on the communication board. Occasionally 
students responded to and corrected their own errors 
after hearing the ther^ist say the word aloud. Thera- 
pists indicated an error or misunderstanding with a 
head shake or by saying “no” or “I don’t understand”. 

When a student omitted an important word, the thera- 
pist sometimes paraphrased the sentence by substitut- 
ing an indeterminate filler for the missing word. For 
example, a therapist paraphrased ‘girl’, ‘make’, ‘in’, 
and ‘pan’ as: “Girl makes something in the pan”. In 
response, the student responded by selecting the 
^propriate word to take the place of “something”. 

Confirmation. Generally, students confirmed the ther- 
apist’s paraphrase gesturally or vocally, but occasion- 
ally select^ ‘yes’ on the communication board. 
Therapists used verbal acknowledgments, head nod- 
ding, and reiterations of the student’s sentence to indi- 
cate their understanding and agreement. 

Thenqiists inhaently offered students the chance to 
confirm or object to their interpretation of each word 
when they echoed the word aloud immediately after 
the student selected it, as they had been instructed to 
do. As well, therapists frequently requested confirma- 
tion from the students explicitly, asking, for example, 
“Is this what you mean?” before or after paraphrasing 
a sentence. 



141 

June 17-22, 1994 



127 



Modelling Partner Interactions 



Other. Ther^ists often gave encouragements, such as 
“You’re doing an excellent job!” or “That’s good”, 
and directives, such as “Ok, we can go on” or “Let’s 
start over hwe”. In addition, therapists asked “Is th«e 
anything else?” or “Are you done with this picture?” 
at almost every picture. It was not always clear, how- 
ever, whetho' the ther^ist was prompting the student 
to say more, or simply inquiring whether to turn the 
page. Some additional commentary by the ther^ist 
occurred due to external distractions. 

FUTURE CONSIDERATIONS 

Preliminary analysis has suggested a number of 
refinements to be considered in subsequent studies. 

Data Collection. A second camera would allow direct 
observation of the symbols selected on the communi- 
cation board. In addition, better sound and lighting 
quality would be helpful for the task of transcription. 

The storybook should be positioned so that the thm- 
pist cannot see the pictures as the student is describing 
them. The knowledge available to the therapist would 
then be more similar to the knowledge available to an 
intelligent AAC system. In the pilot study, the story- 
book was often used as an “extension” of the conunu- 
nication board through pointing gestures. Also, 
several times the therapist tqrpeared to use knowledge 
of the storybook to add information to the descriptions 
that the student had not communicated. 

Design Criteria. In the pilot study, the ther^st played 
a dual role as intoprete- and listener. In order to sepa- 
rate these two roles, a third p^on, unfamiliar with the 
student, may act as the listen^-. The therapist would 
perform the role of an imaginary AAC device, fiteely 
interpreting the communication of the student and 
conveying it to the listener. 

Also, the tho^ist would not be explicitly instructed 
to echo each selection and par^hrase at the end of a 
sentence. Instead, we are interested in the interactive 
dialogue that would occur naturally in this situation. 

Finally, in future studies adult subjects may be used, 
as well as a variety of AAC devices (both electronic 
and manual). In addition, a more familiar environment 
(e.g., home) may be provided for data collection. This 
will help ensure the observation of natural interaction 
patterns. 



terns in the interactive communication that takes place 
between AAC users and their partners. Further studies 
will allow us to define these patterns more accurately 
and completely, forming the basis of a model that can 
be used in the development of future intelligent AAC 
systems. 

REFERENCES 

(1) Kraat, A. (1987) Conmunication Interaction 
Between Aided and Natural Speakers: A State of 
the Art Report. Madison, WI: IPCAS. 

(2) Heim, M. (1990). Communicative skills of non- 
speaking CP-children: A study on interaction. 
Paper presented at the Biennial ISAAC Interna- 
tional Conference on Augmentative and Alterna- 
tive Communication (4th, Stockholm, Sweden, 
August 12-16, 1990). 

(3) Roy, D. M., Panayi, M., Harwin, W. S., & Fawcus, 
R. (1993) Advanced input methods for people 
with cerebral palsy: A vision of the future. Pro- 
ceedings of the 16th Annual RESNA Conference. 
(pp. 99-101). Las Vegas, USA: RESNA. 

(4) van Balkom, H., Kamphuis, H., Demasco, R, & 
Foulds, R. (in preparation). Language technology 
in AAC: Automatic translation of graphic symbols 
into text and/or synthesized speech. 

ACKNOWLEDGEMENTS 

This work has been supported by a Rehabilitation 
Engineering Center Grant from the National Institute 
on Disability and Rehabilitation Research 
(#H133E30010). Additional support has been pro- 
vided by the Nemours Foundation. 

The authors would like to thank the student collabora- 
tors and the staff at HMS School for Children with 
Cabral Palsy, Philadelphia, for their interest and par- 
ticipation. 

Peter Vand^heyden 

Applied Science and Engineering Laboratories 
A. I. duPont Institute 
P.O. Box 269 

Wilmington, DE 19899 USA 
Internet: vandCThe@asel.udel.edu 




CONCLUSIONS 

Observations firom this pilot study reflect the potential 
for discovering common dialogue features and pat- 




RESNA’94 • June 17-22, 1994 



128 



AAC-USER THERAPIST INTERACTIONS: PRELIMINARY LINGUISTIC OBSERVATIONS AND 

IMPLICATIONS FOR COMPANSION 

Kathleen F. McCoy, Wendy M. McKnitt, Denise M. Peischl, 

Christopher A. Pennington, Peter B. Vanderheyden, and Patrick W. Demasco 

Applied Science and Engineering Laboratories 
University of Delaware/A.I. duPont Institute 



ABSTRACT 

Intelligent AAC Systems attempt to provide a com- 
munication system that can interpret input from the 
user in much the same way a familiar listener would. 
The COMPANSION system is a research demonstra- 
tion prototype which “interprets” compressed input 
given by a user of a word based system into a full 
grammatical sentence. In developing a usable system 
from the prototype the needs of the user must be spec- 
ified in well-defined ways. 

This paper reports some preliminary observations 
from an experiment in which word board users inter- 
act with their therapist to tell a story from a picture 
book. The analysis compares the therapist’s output 
with what could be achieved by a system like COM- 
PANSION and discusses the necessary functionality 
for a second generation prototype as well as some of 
the potential difficulties that will be faced. 

BACKGROUND 

In recent years, a number of AAC researchers have 
attempted to develop techniques and systems that 
translate symbol or word input into well formed sen- 
tences (1, 2, 3). Common to the various approaches is 
the ability to inflect words (e.g., verb conjugation) and 
to add function words (e.g., determiners). 

In the COMPANSION technique, a primary emphasis 
was the inclusion of a sophisticated semantic knowl- 
edge base and numerically-based heuristics for rea- 
soning about relative word roles (2, 4, 5). For 
example, the system might take a set of input such as: 
“<apple> <pear> <eat> <john>” and generate “An 
apple and a pear were eaten by John”. Note that in 
order to generate such a sentence the machine had to 
recognize that ‘apple’ and ‘pear’ were the things 
being eaten (recognizing a conjoined theme), and that 
John was doing the eating. In addition, appropriate 
determiners (e.g., “a”) were added (but not to proper 
nouns such as John), and the appropriate passive con- 
struction was used (requiring the past tense form of 
“be” and a past participle ending on the main verb) in 
order to maintain the input order used by the user. 

The COMPANSION approach has been implemented 
as a Lisp-based demonstration running on a Sun 
Workstation. In developing a second generation proto- 
type that will form the basis for a practical system, it 



is necessary to validate the inferencing methods previ- 
ously used and to understand any additional needs for 
a future product. To accomplish this, we need a meth- 
odology for deciding the specific functionality 
needed. Ideally, we would like our system to act like a 
familiar human partner does. Thus, this paper attempts 
to uncover interaction patterns that occur between an 
AAC user and a listener with an emphasis on the types 
of linguistic transformations performed in translating 
word sequences to sentences. 

METHOD 

Pilot data was collected by transcribing videos origi- 
nally recorded by van Balkom. Adolescent students 
with cerebral palsy described pictures in a children’s 
book to their primary speech therapists, using their 
own manual symbol charts. Four such adolescent- 
therapist dyads were videotaped and analyzed. 

Each student was instructed to describe the pictures as 
if telling a story to younger children. The therapist 
was instructed to repeat each word as it was selected 
by the student, paraphrase the sentence when it was 
completed, and then ask the student for confirmation 
that the paraphrased interpretation was correct. A sin- 
gle camera was used to videotape both the student and 
the therapist. Students took between 11 minutes and 
one hour to retell their stories. 

RESULTS 

Some Interactions Consistent with the COMPAN- 
SION Approach: 

Standard Compansion: Some interactions with the 
therapist followed the “standard” operation of the 

compansion system.' 

S: <girl> <make> <in> <pan> <egg> <breakfast> 

T: Girl will make the eggs in the pan for breakfast. 

Here the therapist has added tense, and determiners. 
In addition the plural form of “egg” was chosen. 
Though not indicated by the student, the plural form 
may have been chosen using default knowledge (that 
people generally eat multiple eggs for breakfast) or it 



1. In this and subsequent examples “S” stands for 
the student input and ‘T” the therapist. Words/let- 
ters added by the therapist are in italics. Words of 
particular interest are in bold. 



143 

RESNA’94 • June 17-22, 1994 



129 



AAC User/Therapist Interaction 



may have been the result of extra>linguistic informa- 
tion (e.g„ the picture being described at the time). 
Notice tttat the preposition/or was also included in the 
expanded message. This addition required reasoning 
about the semantics of the input sequence. For exam- 
ple, breakfast was the “reason” for making the eggs 
and should be introduced with a for preposition. 

Word Order Changes: An assumption of the COM- 
PANSION system has been that the words will be 
given to the system in the same order that they should 
be Output in a sentence. However, some of our analy- 
sis reveals that the therapist sometimes did not follow 
the word order initially given by the student. The 
above example falls into this category: the eggs and 
the pan have been switched in the therapist’s output. 
Consider the following example: 

S: <boy> <table> <dusting> <grandmom> <floor> 
<sweep> 

T: Boy is dus ting |f/ie table and the grandmom is 
sweep/ng|f/ie floor. 



Notice that in this instance the student is not following 
a standard subject-verb-object ordering of the words. 
The therapist changes the order to follow standard 
English word order (it is not obvious how to form an 
English sentence while keeping the word order given 
by the student). 

Agent Irference: Another assumption of the COM- 
PANSION system is that a user might omit an agent 
when referring to him/herself. An agent might also be 
omitted if it was obvious from context. This behavior 
was also found in our analysis. Because the story was 
about a boy and a girl, students sometimes did not 
specify an agent, yet it was inferred by the therapist: 

S: <wash> <clothes> 

T: They are wash/ng clothes. 

Verb Inference: Another assumption of the COMPAN- 
.'^TON system is that the main verb may be left out in 
some situations (particularly when the main verb is 
either have or be). We have argued previously that a 
system must have the ability to reason about which 
verb is most appropriate in the given situation. Our 
default rule (i.e., if there is an animate agent and an 
inanimate object, then the verb have should be 
inferred) is consistent with examples found in the 
transcripts. Consider the following where both the 
agent (“they”) and the verb (“have”) have been 
inferred. 

S: <toys> 

T: They have toys. 



RESNA ’94 



Conjunctions: Students sometimes left out conjunc- 
tions in the pilot study: 

S: <boy> <glrl> <made> <bed> <up> <in> <morn- 
ing> 

T: The boy and the girl made up the bed in the 
morning. 

The conjunction could involve the agent role (as 
above) or other semantic roles: 

S: <mom> <heip> <shirt> <shoes> 

T: Mom ’s help/ng with the shirt and the shoes. 

Omitted conjunctions were also observed at the sen- 
tence level: 

S: <girl> <make> <bed> 

<boy> <help> <girl> <bed> 

T: The girl makes up the bed and the boy helps the 
girl make up the bed. 

Possessives: The inference of when a conjunction is 
necessary is complicated by the need to correctly indi- 
cate possessive information. The following example 
contains an inferred possessive. 

S: <boy> <glrl> <girl> <clothes> <my> <mother> 

T: They’re giving their clothes to their mother. 

This example is interesting in that it points out several 
of the difficulties inherent in inferring when a posses- 
sive is needed. Note above there was both a conjunc- 
tion (“boy” and “girl” combined to “they”) and two 
possessives. A possible possessive rule might require 
that if you want a possessive followed by a noun, just 
put the two items next to each other (e.g„ <girl> 
<clothes> for “the girl’s clothes”). Note here <girl> 
<clothes> was translated as "their clothes” as if <girl> 
was now “standing for” the combined agent. How- 
ever, this strategy was not followed for the second 
possessive (the sti^tegy would have resulted in <girl> 
<mother> being used). Rather the student chose the 
first person possessive pronoun, “my”, to indicate the 
recipient in the message. 

It is not clear in the data how much of the therapist’s 
interpretation were influenced by the picture book 
itself. Nonetheless, it raises important questions about 
how to determine when a possessive form is desired. 

Some Interactions Beyond the Scope of the Cur- 
rent COMPANSION Approach: 

Dropped Word (included in interpretation): In some 
instances the therapist did not include words given by 
the student in the interpretation even though they 
often contributed to the intended meaning. Consider: 

S: <two> <table> <in> <table> <dining> <room> 

T: There were things on testable in the dining room. 



June 17-22, 1994 



130 



AAC User/Therapist Interaction 



Notice that <table> occurs twice in the student’s input, 
but only once in the interpretation. In some sense, the 
student’s input is “linguistically” sound. He is saying 
two things about a table (a) there are two things on it, 
and (b) the table is in the dining room. If these two 
assertions were stated as two separate sentences, then 
“table” would occur twice. However, as a single sen- 
tence there is a way to combine the thoughts without 
repeating “table”. Compare this example with the pos- 
sessive case above for an illustration of the difficulty 
in distinguishing this case from that of a possessive. 

Replacing a Word (not included in interpretation): In 
some instances the therapist ignored words selected 
by the student, even though there was no obvious indi- 
cation from the student to ignore the word. 

S: <girl> <help> <clothes> <up> 

T: Girl clothes up. 

She's hanging the clothes up. 

Note that in the above example <help> does not occur 
in the output The example also shows a case where a 
new verb has been inferred (probably from the extra- 
linguistic context). 

More Complicated Verb Inference (Adding or Replace 
ing a Word): In some instances the ther<q)ist inferred a 
verb which was not actually included in the input: 

S: <boy> <girl> <up> <table> <for> <lunch> 

T: ok. They're setting up the table for lunch. 

Dropped Words (not contributing to meaning): In 
some instances the ther^ist dropped words from the 
interpretation: 

S: <girl> <look> <at> <to> <boy> 

T: The girl 's looking at the boy. 

DISCUSSION 

Throughout the study, we observed that much of the 
communication between the student and the therapist 
was not done through the word board. Several of the 
students were quite apt at getting their meaning across 
by a combination of vocalizations, gesturing, and 
pointing (both at the picture book and around the 
room). Such multi-modal communication is beyond 
the ability of most AAC systems available today, but 
would be a fruitful area of future research. However, 
in developing systems, it is important to understand 
which of these interactions are critical to the message 
construction process (e.g., yes/no gestures to confirm 
or correct interpretations), and develop means to sup- 
port them in the system’s interface. 

In the development of “intelligent” AAC systems, it is 
useful and appropriate to look at user-partner interac- 




RESNA^94 • 



tions as a source of input into the analysis and design 
process.The information obtained from such a meth- 
odology can provide guidance and justification for the 
development of system knowledge bases and infer- 
encing mechanisms. We believe that grounding sys- 
tem design in actual human interactions will insure 
that systems developed will be relevant to the needs of 
individuals with disabilities. 

REFERENCES 

[1] HunnicuttS. Bliss symbol-to-speech conversion: 
‘Blisstak’. lournitLoTlhe American Voice I/O So- 
ciety 1986;3. 

[2] McCoy KF, Demasco P, Gong Y, Pennington C, 
Rowe C. Toward a communication device which 
generated sentences. In: Proceedings of the 12th 
AonuaLJ&ESNA Conference . New Orleans, LA: 
RESNA: 1989. 

[3] Reich Peter & Shein F. VOICI: A voice output in- 
telligent communication system. In: Presented at 
the FourthLBJennial ISAAC Conference . 1990. 

[4] Jones M, Demasco P, McCoy K, Pennington C. 
Knowledge representation considerations for a do- 
main independent semantic parser. In: Proceed- 
ings of the 14th AnjiuaLJBESNA Conference . 
Kansas City, MO: RESNA: 1991. 

[5] Demasco PW, McCoy KF. Generating text from 
compressed input: An intelligent interface for peo- 
ple with severe motor impairments. Communica- 
tions of the ACM Mav 1992;35(5):68^78. 

ACKNOWLEDGEMENTS 

This work has been supported by a Rehabilitation 
Engineering Center Grant from the National Institute 
on Disability and Rehabilitation Research 
(#H133E30010). Additional support has been pro- 
vided by the Nemours Foundation. 

The original data collection was performed in collabo- 
ration with Hans van Balkom and Harry Kamphuis, 
Institute of Rehabilitation Research (IRV), Hoens- 
broek. The Netherlands. IRV is an institute for 
research, development and knowledge transfer in the 
field of rehabilitation and handicaps. The authors 
would also like to thank the student collaborators and 
the staff at HMS School for Children with Cerebral 
Palsy, Philadelphia, for their interest and participation. 

Kathleen F. McCoy 

Applied Science and Engineering Laboratories 
1600 Rockland Road, P.O. Box 269 
Wilmington, Delaware 19899 USA 
Internet: mccoy@asel.udel.edu 

145 

June 17-22, 1994 



131 



Changes in Interaction Among AAC Users and their High School Peers 

Cynthia J. Cress 

Meyer Rehabilitation Institute, Omaha, NE 
Ann E. Ratcliff 

Central Michigan University, Mount Pleasant, MI 



ABSTRACT 

This study examined changes in peer communicative 
interactions for five adolescents who used various 
AAC systems. Five repeated single case designs 
were used to address the proposed questions: what 
interactive behaviors are demonstrated by high school 
AAC users and their peers, and what impact does 
peer education have on these interaction patterns? 

The goal of the intervention was to increase 
communicative behaviors by the AAC users and 
nondisabled peers during homeroom class periods by 
educating the nondisabled peers about 
communication. Four communicative behaviors were 
coded on-line by two investigators before and eifter a 
series of peer education discussions. Results 
indicated that various situational influences were 
associated with successful peer interactions, and that 
the effect of peer education alone on communicative 
behaviors was limited outside of these situations. 

BACKGROUND 

Augmentative and alternative communication (AAC) 
system users are at risk for restricted opportunity and 
practice of communicative interaction skills with 
nondisabled partners. Increased educational 
integration has not consistently resulted in more 
and/or better communication between AAC users and 
their age-matched partners (1,4,6). Given the 
changes in social interaction patterns that occur in 
adolescence and the impact of physical and cognitive 
disabilities on social and communicative experience 
(5), adolescents with severe disabilities would be 
expected to demonstrate particular difficulties in 
initiating and maintaining communicative 
interactions. Individual peer education has been a 
successful technique to increase interactions for 
young children (2,3). This study examined the 
impact of group peer instruction on communicative 
interactions for adolescent AAC users. 

RESEARCH QUESTIONS 

1. What are some of the characteristics of 
communication interaction among severely speech 
and cognitively impaired adolescents and their 
nondisabled peers? 

2. Are there changes in the interaction behaviors of 
adolescent AAC users and their nondisabled high 
school peers following group education activities? 



METHOD 

Subjects 

Subjects were five adolescents, 3 female and 2 male 
(15-20 years). All subjects were classified in school 
records as having severe physical and/or cognitive 
disabilities that required AAC intervention. Two 
subjects used electronic voice output devices as their 
primary communication mode, one used a 
combination of sign language, speech, and electronic 
device, and two used a combination of gestures, 
vocalizations, and picture selection. All subjects 
received intervention in the appropriate use of their 
systems throughout this study. 

Procedure 

This research consisted of five repeated single case 
designs. The overall goals of the project were to first 
describe and then increase communicative behaviors 
of the AAC users and nondisabled peers during 15 
minute homeroom class periods by collecting 
baseline data and then educating the nondisabled 
peers about the subject's communication behaviors 
and needs. The education activities included 
discussions related to types of communicative 
behaviors, commonalities and differences in 
communication modes and techniques, and 
motivational factors for communication, and methods 
for improving communication. These activities were 
adaptated from peer education procedures described 
by Cassatt- James (2). Interactive behaviors were 
coded during three baseline sessions and three probe 
sessions within two weeks after intervention. 
Observations were also recorded in a setting in which 
no direct intervention occurred; for two subjects 
(Subjects ID and KS) this was another mainstreamed 
class and for the other three (Subjects SA, CB, and 
TR) this was an off-campus vocational setting. 

Four types of interactive behaviors were coded on- 
line during classroom sessions: attention bids, 
communicative initiations, communicative responses, 
and functional interactive behaviors. Behaviors were 
coded for the subject, and for the teacher and 
classmates when their behaviors were directed 
towards the subject. Percent agreement on behavior 
identification experimental coding was 83-89% 
across subjects. Communicative behaviors were 



RESNA ’94 



graphed by time for each subject and examined for 
, systematic changes between baselines and probes 

146 

June 17-22, 1994 



132 



AAC & Peer Interaction 



RESULTS 

The total numbers of subject, peer, and teacher 
communicative behaviors observed in experimental 
sessions for each of the five subjects are displayed in 
Figure 1. Only two subjects, TR and KS, showed 
increases in communicative behaviors between 
baseline and probe sessions. JD showed a slight 
decline in communicative behaviors, SA 
demonstrated consistently few behaviors, and CB 
demonstrated variable communicative behaviors 
throughout the experimental sessions. Therefore, 
while some subjects demonstrated behavior increases, 
the peer education intervention did not result in 
consistent changes in conununicative behaviors by all 
AAC subjects. In general, communicative behaviors 
changed from baseline to probe sessions for some but 
not all of the subjects, peers, and teachers. While the 
intervention was designed to influence the likelihood 
of situations and attitudes conducive to interaction, 
any discussion of the independent influence of the 
intervention needs to be interpreted in light of other 
potential changes within the classroom itself. The 
circumstances under which changes in behavior did 
occur, and the role of the peer education in those 
changes will be discussed in the following section. 

Communicalivt9 Behaviors in Homeroom 
Class tor Subject SA 




Communicative Behaviors in Homeroom 
Class for Subject JD 



Figure 1: Total Communicative Behaviors 
During Coded Sessions for Five Subjects 



Figure Kev: 

-o- Peer 
-o- Teacher 
Sut^ect 

Communicative Behaviors in Homeroom 
Class for Subject TR 




Communicative Behaviors in Homeroom 
Class for Subject KS 




CommurUcaUve Behaviofs in Homeroom 
Classes tor Subiect CB 



o 

ERIC 





147 



COPY AVAILABLE 



RESNA’94 • June 17-22, 1994 



133 



AAC & Peer Interaction 

DISCUSSION 

Several methodological issues were likely to have 
cdntributed to these results. First, the time 
restrictions imposed by the homeroom class 
environment limited both the length of the 
interventions and the numbers of baseline and probe 
sessions. Also, a relatively brief intervention would 
be expected to have correspondingly subtle impact on 
the behaviors of the participants. On-line 
measurement techniques may not have captured the 
full range of behaviors occurring during the class. 

Some of the most consistent features noted in class 
sessions with higher levels of communicative 
behaviors were situational factors that facilitated 
communicative clarity and opportunity. Factors that 
tended to co-occur with longer or more frequent peer 
interactions included the following: 

Peer -appropriate group activities. Sessions with 
higher levels of student and peer communicative 
behaviors tended to be days on which group activities 
were conducted that involved multiple classmates. 
Sessions in which the subjects initiated activities that 
were less appropriate to peer interests were less 
successful at maintaining communicative interaction. 
Communication routines. Greater initiation by 
subjects and responsiveness of peers and teachers was 
noted in circumstances with well-established 
communicative routines. Many of these routines 
involved greetings, standing jokes between 
classmates, or familiar conversation topics. 

Object and environmental referents. Subjects were 
consistently more successful at obtaining and 
maintaining interactions with peers when those 
interactions related to some object or activity in the 
immediate environment, either available within the 
classroom or brought to class by the subjects. These 
environmental referents tended to facilitate three 
conversational functions: establishing joint context, 
increasing number/length of turns, and encouraging 
peer responses to subject behaviors or initiations. 
Obligatory speaker turns. For some of the subjects, 
interaction was longer and more equally distributed 
between speakers if the activity prompted subject 
responses as well as peer responses. For instance, 
when JD was filling out a crossword puzzle with 
several peers, the situation necessitated a more equal 
exchange of communicative turns. 

Practice of communication techniques in contexts. 
Given the relative unfamiliarity of the nondisabled 
peers with AAC techniques, successful interactions 
tended to involve messages that were directly 
transferrable to ongoing classroom activities. 

Subjects tended not to use the AAC devices in 



homeroom for messages that were frequently used in 
other settings, even if those messages were equally 
applicable to the homeroom. 

In conclusion, brief peer education alone was an 
insufficient intervention for a "quick fix" solution to 
improve interaction. Peers are likely to need 
behavioral modeling within the context to provide 
consistent changes in communicative behaviors. 
However, there may be peers for whom education is 
appropriate and helpful, so we need to examine 
individual differences in responses in future studies. 
Also, students may experience changes in comfort 
level or attitude that are not immediately reflected in 
the observable behaviors coded in this study. Future 
research should incorporate a variety of methods to 
adapt communicative situations to foster interaction. 

REFERENCES 

1. Calculator, S.&Dollaghan,C. (1982). The use of 
communication boards in a residential setting: An 
evaluation. Journal of speech and hearing disorders, 
47, 2SI-2S7. 

2. Cassatt-James, E. L. (1989). The effects of peer 
facilitators on the communicative interactional skills 
of children using communication aids. Unpublished 
dissertation. University of Maryland. 

3. Kraat, A. (1985). Communication interaction 
between natural and aided speakers: An IPCUS 
study report, Toronto: Canadian Rehabiliatioh 
Council for the Disabled. 

4. Light, J. (1988). Interaction involving individuals 
using augmentative and alternative communication 
systems: State of the art and future directions. 
Augmentative & alternative communication, 2, 66-82. 

5. Muuss, R. (1988). Theories of adolescence. New 
York: Random House. 

6. Yoder, D.E. & Kraat, A. (1983). Intervention 
issues in nonspeech communications. In J. Miller, D. 
Yoder, & R. Schiefelbusch (eds.). Contemporary 
issues in language intervention: ASHA report, 27-51, 
Baltimore: ASHA. 

ACKNOWLEDGEMENTS 
Support for this work has been provided in part by 
Grant #MCJ 319152 from the Bureau of Maternal & 
Child Health, US Dept, of Health & Human Services. 

Cynthia J. Cress 

Speech-Language Pathologist 

Meyer Rehabilitation Institute 

600 S. 42nd. Street 

Omaha, NE 68198-5450 USA 

402-559-5754 (office)"or 559-5737 (fax) 




RESNA ’94 • June 17-22, 1994 



. . J'.--'. V 



134 



COMMUNICATIVE EFFICACY OF COMPUTERIZED VIC AND NATURAL LANGUAGE 



Cheryl Goodenough-Trepagnier*, Brent Koeppel** and Susan Stegeman*** 
^School of Audiology and Speech*Language Pathology, Memphis State University 
**Tufts University School of Medicine ••♦Rehabilitation Institute of Michigan 



ABSTRACT 

A Communication Evaluation was 
carried out to determine whether NEWVIC, a version 
of computerized Visual Communication, was more 
effective than natural language for five individuals 
with severe, chronic aphasia. In receptive conditions, 
no group advantage for VIC over either auditory 
comprehension or written language comprehension 
was found. Two individuals demonstrated somewhat 
better success with natural language than with VIC, 
while one individual showed a small disadvantage for 
comprehension of written language as compared to 
comprehension of VIC or speech. One individual 
who participated in an expressive evaluation proved 
to be more successful with speech than VIC. These 
findings underline the importance of individual 
differences in designing alternative communication 
techniques for persons with severe aphasia. 

BACKGROUND 

VIC (Visual Communication) (1, 2, 8) is 
an alternative communication technique for persons 
with severe chronic aphasia. Virtual cards, on which 
content and relational words are represented by 
pictures, are stored in piles according to their 
grammatical category, and are combined according to 
a rudimentary English-like syntax to produce VIC 
messages. 

VIC differs from inert language boards, and 
from most augmentative systems for persons with 
neuromotor disabilities in one primary respect: as 
the user selects lexical items, s/he copies them to a 
message display area where they remain displayed 
and where their spatial order can be easily edited. In 
contrast, the user of a communication board has to 
preplan the temporal order of her/his words or 
symbols and retain a mental representation of their 
order, demands which are usually beyond the 
linguistic and memory capaabilities of persons with 
severe aphasia. 

A number of people with severe expressive 
aphasia have demonstrated superior VIC 
comprehension and production in the laboratory, in 
comparison to natural language, and some of these 
individuals make use of VIC on a daily basis (e.g., 
(3)). However, the majority of VIC learners have 
not reached a level of competence adequate for 



RESNA’94 • 



functional communication. It has been argued that 
this result is unsurprising since individuals with 
aphasia exhibit, in their VIC performance, the same 
deficits that they exhibit in natural language (6). 

RESEARCH QUESTION 

The goal of this evaluation was to 
determine the relative effectiveness of the NEWVIC 
skills acquired by five participants in a study of the 
functional benefit of VIC compared to their natural 
language performance. 

METHOD 

Participants 

Five individuals took part in this 
evaluation. All shared the diagnosis of severe 
nonfluent aphasia from single left hemisphere 
stroke, and had received intensive speech/language 
therapy, until discharged for lack of further progress. 
Scores on the Boston Naming Test and the test 





Gender 


Handed. 


DOB^ 


DOL 


VIC 

Begyn 


DB 


F 


R 


3/29 


3/89 


3/92 


JF 


M 


R 


5/43 


10/90 


7/91 


NG 


M 


AMBI. 


11/33 


1/90 


9/91 


FM 


F 


R 


4/18 


11/90 


7/92 


WS 


M 


R 


8/36 


4/89 


10/91 



Tahlf. 1 : Participants 

of Complex Ideational Material (CIM) from the 
Boston Diagnostic Aphasia Examination (BDAE) 
are offered in Table 2 as indicators of participants' 
approximate level of fluency and auditory 
comprehension at enrollment in the study. All were 
severely expressively impaired, and comprehension 
impairment ranged from severe to moderate. 



Speech Output Auditory Comp. 

/Boston Naming Test) (CIM) 

DB 0 7 

JF 2 0 

NG 0 4 

FM 0 8 

WS 5 9 

Table 2: Speech and Comprehension 

'14B 

June 17-22, 1994 135 



Communicative Efficacyof VIC 



Task 

In Receptive mode the task was picture 
selection. Eleven sets of four pictures each were 
constructed so that syntactic elements needed to be 
processed in order to make the correct selection. For 
example, the sentence for Set 6 indicated that either 
the dollar bill or the comb, was either next to or in, 
the purse. Conditions 1 and 4, contained VIC 
sentences, in Condition 2, the sentence was 
presented auditorily, and in Condition 3, printed 
words replaced the VIC symbols on the monitor. 
Syntactic constructions probed included transitives 
with two and three arguments, intransitives, 
prepositions, adjectives and coordination. No novel 
vocabulary was used. 

In Expressive mode, the task was to 
produce a message describing the event or action 
shown in a picture which was presented along with 
the three others in its set. The participant was told 
that a receiver would try to identify the target picture 
from his message. The receiver stayed on the other 
side of a partition during VIC message production, 
then came over to the computer, read the message 
and made his choice. In the speech condition, the 
receiver listened from the other side of the partition, 
then came over to examine the pictures and select 
the one he thought he had heard described. In both 
cases, the participant decided when the response was 
complete. Only one of the participants, WS, was 
considered capable of carrying out the speech 
condition. 

RESULTS 



across individuals (compared to chance score of 
about 3), and individual averages ranged from 6 to 
10.2. However, comparison of VIC scores to scores 
in either natural language modality fails to show any 
advantage for VIC across the group, or for any 
single individual. Mean VIC score (including the 
initial and final VIC condition, when applicable) 
was 8.3, compared to auditory and written 
comprehension means of 8.5 and 8.2. DB and FM 
scored perfect or near-perfect performance with 
natural language, but did more poorly with VIC. 
WS had more difficulty with wirtten language than 
with either VIC or spoken English. 

Figure 2, below, presents the results for 
WS's receptive performance and his expressive 
performance with VIC and speech. In the expressive 
task, WS laboriously produced adequate spoken 
descriptions which resulted in correct picture choice 
by the naive receiver in all cases, while he succeeded 
in the VIC condition in only 4 of 6 cases. 




Scccftin Ei^ressm 

Figure 2. WS Communication Success 



Participants were generally successful in 
carrying out the receptive task. Scores on all 
conditions averaged 8:4 out of a maximum of 1 1 




H Mean VIC ^Auditory ^Written 



Figure 1. Receptive Communication 

1 



DISCUSSION 

In NEWVIC training, cards displayed both 
picture symbols and text, and investigators named 
the objects and actions, so that approximately 
equivalent experience with speech, text and VIC as 
input were provided, although no explicit training 
was aimed at comprehension or production of natural 
language. This may have contributed to participants’ 
approximately equivalent performance across the 
VIC and natural language modalities. 

All participants were more than 6 months 
post infarct, and all but one were one or more years 
post by the start of NEWVIC training. Accordingly, 
there was little or no expectation of significant 
further recovery of language function. Probes of 
auditory comprehension and written language 
comprehension pre and post NEWVIC training 
showed no significant improvements in either 
modality. Written expression was not probed. 
Unexpectedly, speech fluency was found to improve 
for one individual. The number of relevant lexical 
items produced by WS in the "Cookie Theft" picture 
description task on the BDAE increased dramatically 
jfrom 4 to 12. More strikingly, the number of 



136 



RESNA’94 • June 17-22, 1994 




Communicative Efficacyof VIC 

structural items went from no phrases at all to two 
fully formed sentences. While this improvement 
followed NEWVIC training, the study did not 
provide means for determining whether the 
improvement could be attributed to the NEWVIC 
intervention. 

If indeed VIC training has a positive effect 
on natural language recovery, it may be because 
VIC serves as a "mapping” therapy (7) by bridging 
the gap between conceptual representations of events 
and syntactic formulation. Individuals who benefit 
from VIC presumably have an abilities profile that 
includes deficits corresponding to the compensation 
that VIC provides, and preserved abilities which 
allow them to realize these gains in some modality, 
whether speech, written language, or VIC. In the 
case of EC, an individual with severe expressive 
aphasia and virtually no intentional speech, 
discussed elsewhere (3), training resulted in 
acquisition of an effective VIC communication 
channel, in contrast to absence of spoken or written 
expression. For WS, the VIC "mapping” orthosis, 
coupled with his residual fluency, may have 
benefited voluntary command of spoken language. 

The claim that VIC communication is 
unsuccessful because VIC learners exhibit the same 
deficits in VIC as they do in natural language (6) 
assumes that VIC communication fails if the user 
does not command the rudimentary English-like 
syntax which VIC learners are taught. However, 
analyses of VIC performance both by successful 
VIC learners and by unsuccessful VIC learners (3, 4, 
5) counter that assumption. Utilization of the 
meaning of syntax is in general the source of 
greatest difficulty for all VIC learners with severe 
aphasia, even those who go on to use VIC in daily 
interaction. The ’successful’ VIC user, EC, referred 
to above, uses VIC as a language of ’nouns’ which 
she groups together, and does not use the synt^ in 
which she was trained for representing relational 
meaning (5). While she typically observes the 
default SVO (subject, verb, object) order, this order 
does not necessarily correspond to the meaning 

relations in her messages, since the first person- 
noun in ner message is not necessanly the doer ot 
the action represented by the following verb. Topic- 
comment may be a more accurate description of her 
VIC message organization than subject-verb. 
Investigation of spatial prepositions with several 
VIC learners has shown that a visual analog 
representation, in which pictorial symbols for 
objects are moved in order to depict their relations 
by physical analogy, is more successfully used, and 
perceived as easier and more natural by persons with 
severe aphasia, than are syntactic prepositional 
phrases. It may be that ’standard’ VIC syntax is 
more appropriate as part of a language stimulation 
technique for persons with some prospect of return 
of expressive language, than as an alternative 



communication technique for individuals with severe 
and irremediable impairment. Further development 
and investigation of representational techniques will 
be required to make VIC an optimally useful 
communication tool for persons who have lost the 
capacity for natural language. 

The cumulative results of studies of VIC 
performance by persons with severe aphasia indicates 
that syntax-based VIC is not necessarily the best 
system for all persons with severe chronic aphasia, 
and may be of limited benefit to many of these 
individuals. It is evident that individual profiles of 
preserved abilities and deficits must be characterized, 
and that the system itself, as well as the content and 
emphases of training, must be adapted to fit 
individual needs. 

REFERENCES 

(1) Baker, E. H., Berry, T., Gardner, H., Zurif, E. 

B., Davis, L., and Veroff, A. (1975) Can linguistic 
competence be dissociated from natural language 
functions? 254, 609-619. 

(2) Gardner, H., Zurif, E. B., Berry, T., and Baker, 
E. (1976) Visual communication in aphasia. 
Neuropsychologia^ 14, 275-292. 

(3) Goodenough-Trepagnier, C. (1990). Functional 
communication using VIC. Proceedingsl990 
Conference lEEE/Engineering in Medicine 
and Biology Soc. 12, 3. 1313-1314. 

(4) Goodenough-Trepagnier, C., Askey, D., & 
Koeppel, B. (1992). ACTOR virtual environment 
for communication in aphasia: preliminary results. 
Paper presented at Int. Soc. for Augmentative and 
Alternative Communication, Philadelphia PA. 

(5) Goodenough-Trepagnier, C. (1994) Visual 
Analog Communication. submitted to 
Aphasiology. 

(6) Kraat, A. W. (1990) Augmentative and 

alternative communication: does it have a future in 
aphasia rehabilitation? 4, 321-338. 

(7) Schwartz,M., Saffran, E., Fink, R.& Myers, 
J.(1991). Mapping Therapy. Paper presented at the 
Academy of Aphasia, Rome. 

(8) Steele, R.&Weinrich, M. 1986).Training 
severely impaired aphasics on a computerized visual 
communication sy stem.Proceedings RESNA 
9th Annual Conference. 348-350. 

ACKNOWLEDGMENT 

Partial support for this project was provided 
by NIDRR Grant Number H133G00149. 

Cheryl Goodenough-Trepagnier 

Memphis Speech and Hearing Center, 807 Jefferson 

Memphis, TN 38105 

Voice: 901-678-5800 FAX: 901-525-1282 

151 



RESNA ’94 • June 17-22, 1994 



137 



Improved User Interface for Word Prediction 



Steinar Tyvand, Tor Endestad, Dag Pedersen* and Jan Heim 
SINTEF REHAB, Oslo, Norway 
♦Institute for Special Education, University of Oslo, Norway 



Abstract 

Word pr^ction is a rate enhancement technique for 
coinmumcation and computer access systems. The 
typical user interface of word prediction require 
frequent view shifts between keyboard and display. A 
proposed integrated keyboard-di^lay is developed and 
evaluated theoretically and empirically. The 
theoretical evaluation is not showing signiftcant 
poformance gain for the integrated display, and this 
result will be compared with the empirical 
evaluation. 

Background 

Text based communication (AAC) systems require 
character input through an input device like a 
keyboard. To be effective, the input rate must be 
high enough to produce messages effectively. Many 
users of AAC devices do not manage to type 
characters fast enough to keep the conversation 
balanced. Difiiaont rate enhancement techniques have 
been developed, and one of them is word predictim. 

A word prediction system is suggesting whole words 
to the user, based on the iveviously typed characters. 
TTie predictions may be based on language or word 
statistics, or on more elaborate language models. 

The user intoface is usually consisting of a window 
on rte^ computer or AAC display showing the 
prediction list, and a set of function keys to select 
predicted words. In some systems the prediction list 
is following the cursor for character insertion (1). 
Word prediction on a PC is typically used in 
combination with a word processor. 

In an earlier study (2) we have found that users of 
word prediction do a lot of view shifts between the 
keyboard and the display, and in this study we have 
focused on this user interface side of word prediction. 

Horstmann & Levine (3) has introduced modelling 
techniques to AAC applications. The conclusion 
froro their GOMS-modelling was that word 
prediction is not faster than simple character by 
character input Newell & al. (4) has argued that 
GOMS-modelling of AAC-systems require models of 
disabled users, and should not be based on data from 
expert users. 



Research Question 

The main research question in this study was how to 
improve the user interface in order to mak<». word 
IHediction more effective. The typical word {xediction 
user is looking at the keyboard while typing. But to 
utilise the predicted words, he must look at the 
prediction list every time a new characto- is entered. 
The result is frequent shifts of view between 
keyboard and display. Our hypothesis was that a 
reduced need for Aese shifts would increase typing 
speed and decrease fatigue. 

Method 

A new design for a word prediction user int erface was 
proposed. The design is based on integrating the 
prediction list display into the keyboard. The idea 
was to concentrate the items that the user have to 
focus on during typing and selecting predicted words. 
We fouiid a keybc^ (Comfort™ Keyboard System) 
that is divided into three parts: left hand keys, ri ght 
hand keys and numeric keypad. We develcqied a 
prototype of a small prediction display based on a 
backlight LCD, capable of showirig six lines of 8-10 
mm high characters. A row of prediction selection 
keys was located at both sides of the display. A 
drawing of the display and parts of the keyboard is 
shown in Fig. 1. 

The display has been evaluated through a GOMS- 
analysis (Goals, Operators, Methods and Selection 
rules) (5,6). A GOMS-model is a tool to gimnlatp. 
users behavoir when interacting with a computer 
system. It provides prediction of time to perform 
specified tasks. 

Our GOMS-model is simulatmg three types of users: 
an inexpmenced typist, a mouth stick user, and an 
expoienced typist The GOMS model can predict the 
time spmt on the typing task, and we have compared 
the use of our integrated keyboard-display proUMype 
with the standard connguration with the prediction 
list in a window on the computer display. The 
prracquisites fa- the GOMS analysis was: 

- The words are typed according to word by word 
dictation. 

■ The intended word is shown in the prediction list 
when the two first characters are typed, and is the 
second word in the list 



138 




RESNA ’94 • June 17-22, 1994 



Improved User Int^&ce to Wnd Predicdm 




Fig. 1 : Drawing of integrated keyboard-display. 



- The mouth stick user is typing the characters one 
by one and is looking at the prediction list after each 
character is typed, and is using 2,4 sec on typing a 
character. 

The prototype will be evaluated empirically during 
the ^ng of 1994. This evaluation will give results 
ftom actual use of the integrated keyboard-display. 
An alternating treatment single case design with a 
physically disabled subject, will be used. Typing 
sp^, perceived benefits and any difficulties will be 
recorded. During the evaluation we will get 
information about the relative performance of the 
integrated keyboard-display compared with prediction 
list on the standard display. 

Results 

The GOMS-model predicts correctly that the mouth 
stick user is the slowo* typist, and the expm is the 
faster. In other words, the model is sensitive to 
different user ch^ctoistics. Fig. 2 is showing the 
results for a mouth stick user without prediction, 
with predictions on the integrated keyboard-display 
and with prediction list on the ordinary display unit 
From the GOMS model we get the predictions that 
word prediction is giving rate enhancement for longer 
words (more than 4 characters). The difference 
between the two display alternatives is not 
significant, according. to ^e GOMS-model. The 
difference is only dependent on the diftoent times for 
view shift between teyboard and t^lay. 



The empirical evaluation to be carried out during 
spring 1994, will give a possibility to validate the 
GOMS-results. If the user trials is giving the same 
results, we know that the GOMS-model is 
representing the user in a good way. If we get 
different results, we may be able to improve the 
GOMS-model with upda^ parameter values based 
on experimental data. 

Discussion 

The GOMS-results does not support the hypothesis 
about view shifts being a miyor problem with word 
prediction systems. We do not know if the model is 
sensitive enough to this variable. We have only 
modelled a specific situation whoe word by word is 
typed according to dictation. By varying the differmt 
variables in the GOMS-model, we may be able to 
find out the sensitivity to the variation in user 
charactoistics and scenarios of use. 

The empirical evaluation will also give us a 
possibility to investigate this. We may also find 
otho^ factors that are important when comparing the 
two alternatives. Fatigue may be such a ftictor. 

References 

(1) Swiffin, A.L., Amott, JX., Pickering, J.A. and 
Newell, A.F.. (1987) Adaptive and Predictive 
Techniques in a Communication Prosthesis. 
Augmentative and Alternative Com munication. 
V0I.3/N0.4. 



; 153 



RESNA’94 • June 17-22, 1994 



139 






Improved User IntnfEice Wcnl Piediction 



Mouth stick user 




Word length - number of characters 



Without 

prediction 

Integrated 

keyboard- 

display 

Prediction list on 
ordinary display 



Fig. 2: GOMS-results for typing a word by a mouth-slick usm* 



(2) Tyvand, S., Nilsson, K.B. and Tetzchner, S.v. 
(1992) Evaluation of Linguistic Prediction. IS AAC 
Conferaice Abstracts. Aumentativ e and Alternative 
Communication. VoL8/No.2. 

(3) Hontmann, H.M. and Levine, SP. (1990) 
Modelling of User Performance with Compute 
Access and Augmentative Communication Systems 
for Handicapped People. Augmentative and 
Altanative Communication. VoLd/No.9. 

(4) NeweU, AP, Amott, JX. and WaUw, A. (1992) 
On the validity of User-Modelling in AAC: 
Comments on Horstmann and Levine (1990). 
Augmentative and Alternative C ommunication. 
V0L8/N0.2. 

(5) Card, S., Moran, T. and NeweU, A. (1983) The 
tmchology of Human-C omputer Interaction. 
Hillsdale, NJ, Lawrence Erlbaum Associates. 



(6) Kioas, D£. (1988) Towards a Practical GOMS 
Model Methodology for Usot Interface Design. 
Handbook of Human Computer Interaction. M. 
Hdander, (ed.). Elsevier. 

ACKNOWLEDGEMENTS 

This study was carried out as a pUot project at the 
new Usability Lab in SINTEF, Oslo, which is 
established by a grant form SINTEF strategic 
research funds. 



Steinariyvand 
Senior Research Scientist 
SINTEF REHAB 
P.O.Box 124 Blindem 
N-0314 Osb, Norway 
« +47 2206 73 00 

A « +47 22 06 73 50 



- 154 

RESNA’94 • June 17-22, 1994 



140 



SIG-04 

Drooling 



SWALLOW REMINDER, SALIVA PUMP AND 
CUP WITH FIXED DELIVERY VOLUME 

Michael R Hillman PhD, Andrew R Gammie BA 
Bath Institute Of Medical Engmeering, Bath, United Kingdom 



ABSTRACT 

We present three devices designed to aid people 
with drooling or mouth control problems. The 
swallow reminder regularly emits a tone to remind 
the wearer to swallow. The saliva pump is used to 
suck excess saliva from the mouth. The cup 
delivers a fixed small volume of liquid on each 
tilting of the cup. 

SWALLOW REMINDER 

Background 

Excessive drooling makes social contact dlfHcult 
and special clothes are sometimes necessary. If 
someone can be encouraged to learn to swallow 
regularly they will be able to keep themselves dry 
without recourse to more invasive procedures. We 
have designed a device to assist in the accepted 
therapeutic practice of the use of body-worn 
swallow reminders. 

The swallow reminder has been available for some 
time now (1), and to date 16 have been used. 
Clients are encouraged to swallow as the aid 
bleeps, in the hope that a swallowing reflex will 
develop (eg 2). 

Design 

The aid is constructed using two low voltage 555 
timer ICs, a piezoelectric sounder and an SR44 
1.5V battery. The battery gives over two months’ 
continuous use. The surface mount circuit is 
arranged under the sounder, with the battery 
holder on the reverse side of the board. The 
c.eraU size is 18mm high x 24mm diameter, 
shown in Fig.l. Decorative packaging of the 
device has been left to the user’s choice. The time 
interval between bleeps can be adjusted from 15 
to 120 seconds, with volume control also provided. 




& 



piezo«lectrfc sounder 

pcb with surface 
mount components 

bottery 

volume and time Interval 
cantrol 



Fig. 1. Swallow reminder design 



Eval^tion 

16 aids have been distributed with a variety of 
results. In each case, a speech therapist supervised 
the use of the aid and commented on the results. 

Some clients are now dry and are making 
independent efforts to swallow and to hold their 
lower lip steady. Many clients, however, became 
accustomed to the tone emitted by the aid and in 
the absence of other reminders, and in one case of 
parent motivation, return^ to their previous state. 
Before this accustomisation, a dramatic reduction 
of wetness was usually observed. 

Discussion 

New packaging for the aid is under consideration, 
and we are now supplying them with different 
tones to make accustomisation to the sound more 
difficult. We shall continue to make them 
available at cost price and pursue further publicity. 

SALIVA PUMP 

Background 

People with bulbar palsy, as a result of motor 
neurone disease or other neurological diseases, 
find coping with excessive dribbling both 
problematic and distressing. These problems are 
obviously compounded when mixing socially. 
There is a need for a small discrete suction unit 
which can be used to draw excess saliva and 
mucus away from the mouth. 

Design 

Two prototype devices were constructed to 
investigate different approaches. 

The first device was based around a small (0.4 
atm) vacuum pump (ACL, Alton, UK). The pump 
applied a vacuum to a 150ml bottle which was 
mounted on the outside of a small box 
(65xl20x40mm) which held the pump, batteries 
and other components. Three Nicad cells were 
used which could be recharged by a socket on the 
side of the box. We initially envisaged this unit 
being worn on a belt with a flexible tube to the 
mouth. This was later changed to its being used as 
a hand-held unit with a rigid tube which could be 
placed in the mouth. 



142 



RESNA ’94 • June 17-22, 1994 



Swallow reminder, saliva pump & cup 



The other unit was based around a peristaltic 
pump unit in a box size lS0xl00x60mm. The fluid 
was collected m a ISOml bottle which located 
inside the box, but could be unscrewed from 
below to empty the contents. Power was from a 
single PP3 battery. The device had a shoulder 
strap. A flexible tube could be introduced into the 
mouth. 

Evaluation 

These devices were evaluated through discussion 
with therapists and short trials with potential 
users. It was soon clear that these devices did not 
provide the required solution, but provided a 
valuable starting point for discussion, llie size was 
the main problem, as it was felt that a unit would 
need to be small enough to flt in either a pocket or 
handbag. One of the main constraints on size is the 
size of the fluid collecting bottle. The bottle used 
had excess capacity for the likely usage and so 
considerable saving of space could be made in this 
area. 

A rigid tube was preferred, though this should 
have a soft end so as not to injure the inside of the 
mouth. The rate and power of suction was the 
other major area of concern, though it should be 
noted that during the trials the vacuum pump was 
found to be not working properly. 

An area which needs to be carefully considered is 
that of the fluid reservoir, ease of cleaning and 
hygiene. It is vital that the fluid should ea^y be 
drained. It is also important that the user should 
be able to see how full the bottle is, though an 
audible warning when the bottle is full should also 
be incorporated. A particular problem with the 
vacuum unit is the n^ to reliably seal the bottle 
when it is screwed into the unit, as any air leaks 
reduce the effectiveness of the suction. 

PgYfflQpmcat 

The unit has been redesigned to meet the 
comments received during the evaluation. A 
vacuum unit is used which applies suction to a 
30ml bottle. The bottle is hous^ within the box 
(7SxlO0x4Omm) which is cut away to allow the 
level of fluid to be observed and to allow the 
bottle to be readily unscrewed. A rigid tube is 
mounted on the outside of the box which folds 
away for carrying. 



Discussion 

We believe we have determined the appropriate 
specification for a saliva pump and have built a 
unit which meets the requirements. This will 
shortly be evaluated before more detailed 
development takes place. 




LIMITED VOLUME CUP 
Background 

Many people cannot drink from a standard cup for 
various reasons. This may be due to poor hand 
control or inability to swallow readily (dysphagia). 
Various cups are available with special handles 
and with lids which restrict the flow rate. These 
have two particular problems. The flrst is that they 
are unaesthetic and look more suited for babies 
tl^ adults. Secondly it is more appropriate for 
many conditions to restrict the volume of fluid on 
each mouthful, rather than the rate of flow of 
fluid. 



The basic design uses an inner cup within an outer 
as illustrated. A reservoir is formed between the 
bottom of the inner cup and the outer. Fluid flows 
from the inner cup to Ae reservoir throng a small 
hole. When the cup is brought to the mouth and 
tUted the fluid flows up a chunel to the mouth. At 
the same time the level of the fluid falls below the 
hole so that no more fluid can flow into the 
reservoir. 

A batch of flve cups were made up to the above 
design, using a standard plastic ^cnie cup as the 
outer cup. This was an aesthetically attractive 
design. The inner cup was vacuum moulded and 
the cap was machined from solid. 



157 



June 17-22t 1994 



RESNA ’94 



9 



143 



References 



Swallow reminder, saliva pump & cup 




Rg. 3. Limited volume cup. 

EvalMtipp 

The general concept was praised and the cup was 
effective. There were however two comments 
which repeatedly came back. One was that the cup 
needed to be tilted a long way for the fluid to 
come out. This involved tiltii^ the head back 
which exacerbated swallowing problems. Secondly 
the nose tended to come into contact with the cap 
of the inner cup. Some users would have preferred 
a two handled design. 



1. "Electronic devices for speech therapists" 
Gammie AR, RESNA 13th Annual Conference, 
Washington, 1990, p.141-142. 

2. "The Meldreth dribble control project 
reassessed" 

Jones PR, Child care health and development, 
82(8), 65-75. 



Dr Michael Hillman, 

Design Engineer, 

Bath Institute of Medical Engineering, 
Wolfson Centre, 

Royal United Hospital, 

Bath BAl 3NG 
United Kingdom 



Development 

In order to decrease the amount which the cup 
needs to be tilted the angle of the cup should be as 
shallow as possible. The degree of shallowness is 
limited however by considerations of both stability 
and aesthetics. In order to reach the best 
compromise the shape of the cup became a tilted 
cone. Since no commercially available cup is 
available to this shape both the inner and outer 
have to be moulded using the vacuum forming 
process. This gives a less attractive finish to the 
cup. However possible volume production may 
allow injection moulding. This new design of cup 
is shortly to undergo evaluation. 

Bisoissipn 

The concept of a limited volume cup has received 
wide interest, and evaluations have ^own that the 
basic design is sound. A redesign to meet the 
detailed f^back is awaiting evaluation. If this is 

f fositive we hope to investigate commercial outlets 
or the design. 




144 



RESNA ’94 



t 



June 17-22, 1994 



SIG-05 

Quantitative Functional Assessment 




OBJECTIVE EVALUATION SCALES FOR THE QUANTITATIVE ASSESSMENT OF 
WALKING FUNCTION IN STROKE PATIENTS 

Antonina G. Starita*, Sandro Bortone*, Antonio Bacciglieri**, Nino Basaglia** 

•Dept, of Informatics • University of Pisa 
C.so Italia n.40 - 56 1 25 Pisa - ITALIA 
••Unit of Functional Rehabilitation • S.Giorgio Hospital- Ferrara - ITALIA 



ABSTRACT 

In this paper it is shown a quantitative method 
of diagnostic and prognostic evaluation of the 
hemiplegic patient in the acute phase, based on 
the combined application of the objective 
evaluation scales (neurologic, sensoiy/rootor, 
huictional), trying to establish a quantitative 
classification of the hemiplegic patients and to 
identify the more sig^cant prognostic 
variables, in order to provide predictive 
elements of the recovery of the walking 
function. 

BACKGROUND 

The recovery of the walking function 
represents one of the main objectives of the 
rehabilitation treatment of stroke patients and, 
for this reason, a prognostic evaluation of such 
a recovery is of great interest. To this aim, it is 
very important that the prognostic methods 
should be, as possible, quantitative and 
repetible and this can be obtained if people 
acting in the rehabilitation field have at their 
disposal tools for a precise evaluation 
according to the different clinical "dimension" 
of disable subjects. 

If the classification of impairments and 
disabilities has the previous mentioned 
carachteristics, the motor and functional 
evaluation of the patient provides several 
advantages to the clinician; 

1, It allows to assign the patient to a 
predefined fhnetional class on which it could 
bo possible to plan a correct rehabilitative 
prqject, to choose realistic objectives fbr the 
treatment, to plan the recoveiy process and to 
quantify the results of the treatment. 

2 , It allows to obtain numeric data, necessaiy 
to peifbrm quality controls and, consequently, 
to manage in a more accurate way the 
rchabilitaUve unit through information on the 
necessary resources in dependence of the 
degree of disability of the patients, 

3, It allows to go in deep with the classification 
of motor and functional recovery, by taking 
into account many clinical variables, and by 
using numerical techniques to process them. 



In literature, there are several approaches to 
the predictivity of the functional and walking 
outcomes, even if all of them can be ascribed 
to one of three different methodologies. 
Among these we have adopted the one that 
concerns the study of reliable prognostic 
criteria using objective evaluation scales 
applied to the quantification of impairments ' 
and disabilities. In this frame we have 
considered as particularly important the works 
of Wade and coll. (1] that have used the 
Barthel functional e>^uation index with 
prognostic aims and Dettman and coll. [2] that 
using the Fugl-Meyer sensory/motor 
evaluation test were able to establish good 
prognostic correlations among the items 
evaluating the sitting and upright posture and 
the walking ability at the patient discharge. 

METHOD 

The study concerns 33 patients affected by 
stroke on an ischemic or emorrhagic vascular 
basis, sequentially admitted near the acute unit 
of the S.Anna Hospital and, successively, in 
the Unit for the Functional Recoveiy and Re- 
education of the S.Giorgio Hospital in Ferrara 
during the period from February 1991 to 
November 1992. All those patients that at the 
stroke onset presented secondary disabilities 
have been excluded from this study. The 
informed consensus to be submitted to the 
clinical evaluations, it has been obtained by all 
the patients . 

All the patients have been evaluated by a 
physiatrist inside the 15th day firom the stroke 
onset near the first recoveiy unit (neurological 
unit) and to each of them the following tests 
have been applied; 

1, Fritz- Werner (FW) test modified by Mathew 

13]; 

2, Fugl-Meyer (FM) test modified by 
Lindmark and Hamrin [4|; 

3, Barthel index (BI) modified by Wade [Ij; 

All the patients have been submitted to 
rehabilitative treatment, either near the acute 
recovery units and in the rehabilitation unit. 

At the discharge the walking function has 
been classified according to three functional 



J 60 

RE@NA’94 • June 17-22, 1994 



146 



Objective evaluation in stroke 

levels ( to which the labels C, A and N have 
been assigned): 1) Autonomous Walking, 2) 
Walking with Devices (ankle foot orthosis or 
crutch); 3) No Walking. 



in Table 1 for the 5 classes case and in Table 2 
for the 6 classes one. 

Plot of Clusters 



RESULTS 

Figures 1 and 2 (5 and 6 classes respectively) 
represent the results of the cluster analysis 
connecting the FM and BI tests (y and x axis 
respectively). In Figure 1 two main 
aggregations can be individuated; 

1) The first is characterised by a group of 
patients with low score in BI and high degree 
of variability of the FM score (c class): 

2) The second (b class) and the third (a class) 
demonstrate less variability in FM with a 
substantial increase in tlie BI test score. 

Class d is restricted to just one patient and it 
appears to be near to normal values. Class e, 
also restricted to just one patient (probably for 
the anomalous value in the FW test not showed 
in the bidimensional representation), is at an 
intermediate level between a and b classes. 

PlotorQustera 




obtained with the cluster 
algorithm on the Barthel and the 
Fugl-Meyer scores. 

In Figure 2 a further application of the same 
algorithm on 6 classes ilows to subdivide the 
c class in other two subclasses (j and k). In 
other terms the algorithm allows to identify 
two well distinct classes that are characterised 
by a low score in the BI index. 

The classes obtained in this way have been 
compared with the outcome values for the 
walking ability, previously illustrated in the 
methods section; the results have been reported 



240 

210 

160 

s 

H 150 
120 
90 
60 

TB 

Figure 2 - Cluster algorithm applied on 
6 classes. 






Autonomo 


Walking 


No 




us 

Walking 


with 

Aids 


Walking 




C 


A 


N 


Class a 


2 


1 


0 


Class b 


3 


0 


0 


Class c 


9 


6 


10 


Class d 


1 


0 


0 


Class e 


1 


0 


0 


Table 1 - Walking 


outcome frequencies 


respect to the 5 classes. 






Autonomo 


Walking 


No 




us 

Walking 


with 

Aids 


Walking 




C 


A 


N 


Class a 


2 


1 


0 


Class b 


3 


0 


0 


Class j 


7 


4 


5 


Class k 


2 


2 


5 


Class d 


1 


0 


0 


Class e 


1 


0 


0 



Table 2 - Walking outcome frequencies 
respect to the 6 classes. 



Once it was established that by scores it is 
possible to obtain the clustering of data it was 
tried to identify the existence and the 
significance of the possible correlations. Table 
3 shows the results of the Kruskal-Wallis test 
performed on the total scores of the three tests 
respect to the three levels of the walking 
outcome. 



c 161 

RESNA’94 • June 17-22, 1994 



147 



Objective evaluation in stroke 







FM 


FW 


BI 


Group 


A 


13.14 


10.92 


12.92 


Group 


C 


22.68 


22.59 


20.81 


Group 


N 


10.60 


12.30 


13.75 



Level of 4.007E-3 4.90E-3 0.081 

Significance 

Table 3 - Results of the 

application of tlie Kruskal-Wallis 
test to the total score of the three 
tests respect to the three walking 
levels. 

The same kind of analysis has been repeated 
on the single items of FM: the most significant 
results in tenns of predictivity have been 
obtained by the sections evaluating mobility 
(AMIP), balance (ABIP) and tactile sensibility 
of the inferior limb (ATSIP) (see Table 4). 

AMDP ABDP ATSDP 
Group A 14.92 12.85 09.92 

Group C 21.93 22.81 23.03 

Group N 10.55 10.60 12.30 

Level of 0.011 2.97E-3 5.67E-4 

Significance 

Table 4 - Results of the 

application of the Kruskal-Wallis 
test to the mobility, balance and 
tactile sensibility items of the 
Fugl-Meyer test. 

DISCUSSION 

The clusters obtained by the total scores of the 
three tests, based only on the intrinsic property 
of the adopted metrics, are an interesting result 
from a methodological point of view, but by 
themself they would have a poor clinic^ 
applicability if are not compared with the 
reliable parameters of the rehabilitative 
process. As it can be noted from Table 1, for 
the classes a, b d and e which collect in all the 
24.25% of the sample population, the 
probability to walk again at the discharge is 
very high (7 patients classified with 
Autonomous Walking and 1 patient with 
Walking with Devices), so they could represent 
a prognostic criteria for the clinician. Inside 
class c which collect the 74.75% of the sample 
population the situation is less clear: in fact 15 
patients on 25 belong to the hmctional classes 
C or A and 10 belong to the N class. The 
further classiEcation of the c class in the two 
subclasses j and k (16 and 9 patients 
respectively) is able to explain better the 



prognostic Judgement. In fact in class j 11 
patients on 16 belong to the C or A categories, 
that qualify walking, while in class k only 4 
patients on 9 belong to such categories. So 
extimating from the observed frequencies the 
probability levels it could be possible to say 
that the probability for a patient of the class j 
to walk again is more then 1.5 times the 
probability of a patient of class k. 

The maximum level of significance respect to 
the ranks created by the Kruskal-Wallis test 
was reached by the FM test. Also the FW test 
has obtained a good significance level but it 
appears less sensible to discriminate patients 
who will walk with aids by those who will not 
walk at all. The BI index applied in the first 
stage of the recovery has no predictive value so 
as the superior limb section score of FM. 

In particular, looking at the single items of the 
Fugl-Meyer we can note that the mobility item 
even if didn't reach a high level of significance 
it is the best predictor of the future walking 
ability, while the balance item reached a very 
good level of significance with still good 
predictive capabilities. Even if the sample 
population in the future will be expanded to 
acquire more statistic soundness our results 
demonstrate the importance of a multi-factor 
approach to the study of this pathology aiming 
at the assessment of the correct rehabilitative 
treatment. 

REFERENCES 

[1] Wade D.T.et al.: Predicting Barthel ADL 
score at 6 months after acute stroke. Arch. 
Phys. Med. Rehabil., 64, p. 24-28, 1983. 

[2] Dettman M.A., et al.: Relationship among 
walking performance, postural stability and 
functional assessment of hemiplegic patients. 
Am. J. Phys. Med., 66, p. 77-90, 1987. 

[3] Lindmark B., Hamrin E.: Evaluation of 
functional capacity after stroke as a basis for 
active intervention, Scand. J. Reliab. Med., 20, 
p. 103 - 109, 1988. 

[4] Matew N.T., et al.: Double-blind 

evaluation of glycerol therapy in acute cerebral 
infarction. Lancet, 2, p. 1327-1329, 1972. 
ACKNOWLEDGEMENTS 

This work has been supported by the AIM 
Project of the C.E.C., CAMARC II. 

Antonina Starita, Assoc. Prof of Bioengin. 
Dept, of Informatics, University of Pisa 
C.so Italia,40 - 56125 Pisa - ITALIA 



148 



1-62 

RESNA ’94 • June 17-22, 1994 



PRELIMINARY DESIGN OF AN OPTIMIZED JOGGING PROSTHESIS 



Denis J. DiAngelo 

Department of Biomedical Engineering, 
University of Tennessee - Memphis 



ABSTRACT 

Despite many amputees’ interest to participate in 
athletic activities, no available lower-limb prosthesis 
allows an above-knee amputee to jog safely in a 
natural fashion. The objective is to design an 
above-knee prosthesis capable of replicating the 
dynamic motion of the normal lower extremity 
during jogging in sports activities. A mathematical 
model representing a novel design was developed to 
facilitate complete simulation of the stance and 
swing phases. The two main components were the 
Knee Control Unit (KCU) and the Swing Phase 
Mechanism (SPM). The KCU provided controlled 
knee flexion during stance, while absorbing the 
impact loads; it also generated some forward and 
vertical motion. The SPM permitted the amputee to 
swing the prosthetic leg freely during non-support 
periods. Information obtained from the simulation 
study was employed in the design of a prototype 
prosthesis. The unique features included a more 
natural, one-to-one stepping pattern, thus enabling 
an above- knee amputee to participate in sports that 
involved jogging. 

BACKGROUND 

Limitations of Existing Designs 
Amputees who attempt to jog with their 
conventional prosthesis acquire an unorthodox and 
asymmetric gait pattern to vault over the 
straightened prosthetic leg; they take two short 
quick steps with the natural limb for each step on 
the prosthetic leg. Two steps on the non-amputated 
limb are required to provide sufficient time to swing 
the prosthesis through to complete extension prior to 
heel contact. Vaulting over the prosthetic leg 
reduces the drop in the stabilizing hip joint and 
increases pelvic rotation; these actions are great 
penalties as compared to normal motions. Knee 
flexion during stance is not attainable with 
conventional knee units and results in considerably 
higher reaction forces, which can cause 
degeneration of the residual hip joint and a 
shortened life for the prosthesis. Conventional 
swing control devices do not provide sufficient 
control during swing activities, as they are designed 
to operate over a range of walking speeds and do 

RESNA’94 • 



not cater to the increased knee angular velocities 
associated with jogging (3,5). Lastly, there are no 
known regulatory safety standards which dictate the 
acceptable design loads for lower extremity sports 
prostheses. Consequendy, it is doubtful that 

conventional components can provide safe 

operation without any mechanical failure, if used for 
jogging purposes. 

The Seattle foot and Flex foot were designed for 
active below-knee amputees and consisted of a 
plastic inner leaf spring that absorbed energy on 
heel strike and returned it in the form of lift in an 
attempt to emulate the toe-off thrust of a non- 
amputee jogger (2). However, personal comments 
from an above-knee amputee who tried jogging with 
the different feet indicated that they provided no 
appreciable aid in permitting a more natural jogging 
stride. 

The Terry Fox Jogging prosthesis was designed to 
alleviate the asymmetric jogging pattern inherent 
with the use of conventional prostheses. The 
prosthesis was comprehensively assessed in the gait 
laboratory and was not found to be an optimal 
design (3). Nevertheless, the information gained 
from this study was deemed useful in the design of 
the proposed physiological prosthesis. 

INTRODUCTION 

Need 

The number of young amputees in North America 
continues to grow, due to cancer and traumatic 
events. Many of these individuals were active prior 
to amputation and wish to continue an active sports 
life. Jogging, which is the basis for many sports, 
remains the most difficult activity to achieve and 
causes the most discomfort (1). Commercially 
available walking prostheses can not be easily 
adapted for jogging. Hence, there is a growing 
market for a sports-oriented prosthesis for above 
knee amputees that will allow a balanced, 
physiological prosthetic to natural leg jogging 
stride. 

Objectives and Design Criteria 

This paper presents the preliminary design of an 

163 

June 17-22, 1994 



149 



OPTIMIZED JOGGING PROSTHESIS 



above-knee prosthesis capable of reproducing the 
dynamic motion of the normal leg during jogging. 
The necessary design criteria were established from 
non-amputee jogging patterns (4). The design goals 
for the stance phase are to absorb the large ground 
reaction forces at heel contact, provide self 
stabilizing knee flexion, and generate some forward 
and vertical motion. The prosthesis should also 
provide reliable support during stance. To achieve 
these objectives, about 20® of knee flexion was 
desired at heel contact. Such controlled knee 
flexion lowers the center of mass of the body, thus 
eliminating the need to vault over the prosthesis. 
During the swing phase, the prosthesis must first be 
accelerated and then undergo controlled 
deceleration prior to heel contact. Lastly, some 
amount of knee flexion at the end of swing is 
necessary to stabilize the joint. 

DESIGN ANALYSIS 

Mathematical Description of Prosthesis 
A mathematical model representing a novel design 
of a jogging prosthesis was developed to facilitate 
complete simulation of the prosthesis for both 
stance and swing phase operations. The 
components of the optimized jogging prosthesis are 
shown in figure 1(a). A conventional socket and 
prosthetic foot were incorporated into the design. 
The prosthesis was modeled as a three link system, 
as shown in figure 1(b). Link 1 represented the 
thigh and socket, and link 2 represented the shank 
tube and prosthetic foot. The Swing Phase 
Mechanism (SPM) was located in link 2 and 
consisted of two buffer springs fastened to the shank 
tube at the end-points of travel of the slider joint. 
The third link represented the Knee Control Unit 
(KCU) and consisted of a large compression spring 
enclosed at both ends by parallel plates, connected 
together by a shaft passing through the center of the 
spring. The position of the upper end plate was 
adjusted to create a preload force on the coil spring 
and served to regulate the onset of knee flexion 
during stance. Once the impact force exceeded the 
preload force the spring compressed, causing the 
knee to flex. Links 2 and 3 were coupled during the 
stance phase and decoupled during swing using a 
latching mechanism located inside the shank tube. 
The timing and triggering of this device was 
established from the kinematic analysis of non- 
amputee jogging patterns (5). The latching 
mechanism returned the knee to a flexed position 
prior to heel contact, which was beneficial in 
initiating absorption of the large reaction forces 
encountered at heel strike. 




Figure 1: (a) Schematic of the optimized jogging 
prosthesis showing the KCU and SPM. (b) 
Simplified three link model of prosthesis. 

The Lagrangian formulation was employed to 
derive the equations of motion of the three link 
system and used to design the KCU and SPM. 
Similar procedures have been implemented for 
simulating human locomotion (4). 

RESULTS 

Design of KCU 

The KCU consisted of a spring unit positioned 
posterior to a single axis knee joint and attached 
between the base of the socket and a slider joint on 
the shank tube. The spring size and position were 
determined by an optimization program which 
minimized the forces acting on it, as well as its 
weight. The force in the unit was calculated for 
prescribed knee moment and angle patterns taken 
from non-amputee joggers (5). The resultant knee 
angular pattern for a coil spring with a stiffness of 
45.5 kN/m is shown in figure 2(a). The resultant 
motion compared well with the desired motion of a 
non-amputee jogger. Note that with conventional 
prostheses knee flexion is prohibited during stance, 
as the knee must maintain a hyper extended locked 
configuration. 

Design of SPM 

The main design goal of the SPM was to have the 
prosthesis simulate the swinging motion of the 
lower leg of non-amputee joggers. By subjecting 
the three link system to a prescribed hip and thigh 
trajectory, the resultant motion was determined with 
a numerical differentiation technique. The design 
entailed determining the location and stiffness 
values of the buffer springs; this dictated the range 
and rate of swing for efficient jogging. Figure 2(b) 
reveals a close concurrence exists between the 
resultant and desired angular knee patterns. 




RESNA’94 • June 17-22, 1994 



150 



OPTIMIZED JOGGING PROSTHESIS 




Figure 2: Computational results from the simulation 
study of the prosthesis for (a) stance and (b) swing. 
In both phases, the resultant knee angle patterns 
show close concurrence with the desired non- 
amputee patterns. 

Design of Latch Mechanism 
A spring-loaded latch mechanism was positioned 
inside the shank tube and used to couple links 2 and 
3 during stance and decouple them during swing. 
The rotational energy in the swinging links was used 
to lock the latch mechanism prior to heel contact, 
while the ground reaction load transmitted during 
stance was used to unlock the latch. 

Prototype Design 

A photograph of the prototype prosthesis is pictured 
in figure 3. The weight-bearing, knee-flexing 
feature of the KCU is demonstrated in the figure. 

DISCUSSION AND CONCLUDING REMARKS 




Figure 3: Prototpye of optimized jogging prosthesis. 

with conventional walking prostheses, the knee 
remains locked in a hyper extended position, 
thereby providing minimal resistance against, the 
large impact loads. 3) As soon as weight was 
removed from the prosthesis, a latch mechanism 
disconnected the KCU from the system and allowed 
the leg to rotate freely during the swing phase. A 
set of buffer springs established the limits of flexion 
and extension during swing and acted to reduce the 
total swing phase time. The upper buffer spring 
worked with the locking mechanism to provide 20° 
knee flexion to prior to heel contact. This closely 
simulated the normal jogging motion. These 
features of the prosthesis provided a more accurate 
simulation of the natural motion of the leg, and 
allowed for a symmetrical stepping pattern. 

REFERENCES 



The optimized prosthesis consisted of three major 
innovations. 1) It enabled the jogger to initiate heel 
contact with the knee in a flexed position, which 
helped to absorb the large impact forces. The knee 
continued to flex as the stance phase progressed, 
thus reducing pelvic rotation. Conventional 
prostheses have to be hyper extended at the end of 
swing so that they will be in a locked-knee 
configuration during the stance phase, thereby 
requiring the user to vault over the prosthesis into 
the toe-off position. This requirement also 
contributed to the increased reaction loads and 
excessive pelvic rotation. 2) As the knee flexed 
during stance, energy was absorbed in the main coil 
spring, and the KCU was so designed that the 
appropriate amount of resistance was provided 
throughout stance and stabilized the lower 
extremity. Most of the energy was returned to the 
jogger during the latter half of stance, which helped 
propel the jogger forward and upward. Note 



1. Kegal, B., Webster, J.C. and E.M. Burgess. 
(1980) Recreation Activities of Lower Extremity 
Amputees: A Survey. Arch Phys Med Rehab, 61: 
258-264. 

2. Enoka, R.M., Miller, D.I. and E.M. Burgess. 
(1982) Below-Knee Amputee Running Gait. Am J 
Phys Med, 61: 66-84. 

3. DLAngelo, D.J., Winter, D.A., Ghista, D.N. and 
W.R. Newcombe. (1989) Performance assessment 
of the Terry Fox jogging prosthesis for above-knee 
amputees. J Biomech, 22: 543-558. 

4. Chow, C.K. and D.H. Jacobson. (1971) Studies in 
human locomotion via optimal programming. Math 
Biosci, 10: 239-306. 

5. D.A. Winter. (1983) Moments of force and 
mechanical power in jogging. J Biomech, 16: 91-97. 

Mailing Address: D.J. DiAngelo, Department of 
Biomedical Engineering, UT - Memphis, Suite 801, 
899 Madison Avenue, Memphis, TN 38163. 



RESNA’94 • June 17-22, 1994 



151 



GATT SIMULATION AND THE DESIGN OF A PAEDIATRIC ABOVE-KNEE 
ENDOSKELETAL RUNNING PROSTHESIS 



Alan R. Morris, M.A.Sc.*^^, Stephen Naumann, Ph.D/^, and William L. Cleghom, Ph.D.t 
The Hugh MacMillan Rehabilitation Centre* 

Department of Mechanical Engineering^ / Institute of Biomedical Engineering^, University of Toronto 

Toronto, Canada 



Abstract 

A unique approach combining quantitative measurements 
of human gait, analytical investigation and specification 
of mechanical components, and computer simulation 
techniques, was undertaken to design and eveduate a new 
prototype paediatric above-knee endoskeletal running 
prosthesis. To overcome the performance restrictions of 
conventional children's prostheses, prosthetic components 
were designed to approximate kinematic and kinetic gait 
performance of able-bodied subjects by improving knee 
control and the body's centre-of-mass (COM) 
displacement profile. Experimental walking and running 
subjects were used to provide gait data for component 
design and the computer simulation of gait. Three 
components were proposed to improve ruiming for 
children with an above-knee amputation (AKA): a four- 
bar linkage knee joint, a double-acting knee damper, and 
a telescoping shank spring-damper. For the first time, 
computer simulation was used for preliminary testing and 
evaluation of a complex above-lmee prosthesis (AKP) 
over both the swing and support phases of gait. 
Kinematic simulation results indicated that the prototype 
prosthesis demonstrated both performance improvements 
and weaknesses compared widi a conventional children's 
prosthesis. 

Background 

Children's Gait 

An estimated 10,000 juveniles with an AKA across North 
America [1] are restricted to using simple functioning 
lower-limb prosthetic devices \^^ch place limitations on 
the type and intensity of physical activities they can 
undertake. On a daily basis the "average" child will 
perform activities of different types and intensities than 
the "average" adult. These activities might include 
running, jumping, and kicking a ball. 

Child Amputee Gait 

Able-bodied children differ from adults in terms of: i) 
stride length, ii) cadence, iii) swing time, iv) support 
time, and v) "comfortable" walking and running velocities 
[2, 3]. 



is more critical than in adults due to relatively weak hip 
extensors. The children's prosthetic limb absorbs less 
shock at heel contact since ankle and knee joints cannot 
effectively dissipate the force. At toe-off, die user must 
compensate for this through increased hip/thigh muscular 
force over an increased hip flexion/extension range. 

Running and Individuals with an AKA 
Individuals with an AKA inexperienced at running lend 
to use a hop-skip method; this method (incorporating an 
extra hop) utilizes excessive hip and knee flexion to 
enable hip clearance over the supporting prosthesis [6]. 
Individuals with an AKA require knee extension during 
support, generate larger ground reaction forces and have 
severe problems in generating the propulsive impulse 
needed to maintain equal step lengths between natural 
and prosthetic limbs [7]. 

Above-Knee Prosthetic Advances 
Present AKP prescriptions for children are limited to 
simple units with friction-based control of a single-axis 
knee; most designs have been scaled from adult units. 
These units are inadequate to address running. A number 
of promising knee controllers used in adult AKP have not 
been adapted to children's units: polycentric knee 

linkages, hydraulic and pneumatic controllers. 

Computer Simulation of Normal and Prosthetic Gait 
Most computer simulations of gait are based on applying 
experimentally-recorded or subjectively-specified forces 
to a simplified model of the human anatomy. The 
validity of these approaches is based on achieving 
satisfactory kinematic and kinetic results. Swing phase 
simulations [8] are typically planar, driven by 
experimental j oint moments, >\frile support phase 
simulations have applied heuristically chosen (or 'fudged') 
joint moments to planar [9] and three-dimensional [10] 
models to achieve gait resembling the experimental case. 

Simulations of AKP gait have only been achieved for the 
swing phase [11, 12] by adding polycentric knees and 
various controllers to the typical able-bodied gait model. 

Statement of Problem 



Limited study of child amputee gait has been performed. 
The group of Hoy, Zemicke, and Whiting [4, 5] 
performed studies of five children with AKA donning 
conventional prostheses. Compared to able-bodied 
coimterparts, children with AKA walking showed 
differences in a large number of gait parameters. 

Hoy, Whiting, and Zemicke [4] assert that knee stability 
at initial foot-ground contact of children with amputations 

RESNA’94 • 



• Conventional AKPs for children limit physical 
activity such as running. Prostheses should 
control: support phase knee stability, swing phase 
knee rotation, and impact energy storage and 
return. 

• Typically, prosthetic design is a trial-and-error 
approach which is both costly and time consuming. 




June 17-22, 1994 



152 



GAIT SIMULATION and PROSTHESIS DESIGN 



O 

ERIC 



Rationale 

The objective of this research is twofold: 1) to develop a 
conceptual design for a running prosthesis for children 
with an AKA; and 2) to develop a computerized 
simulation of gait to enable preliminary testing and 
evaluation of prosthetic design ideas. 

Design / Development 

Since stability during support has been a major 
consideration, a polycentric laiee was proposed to provide 
stability during heel-contact, while enabling the initiation 
of knee flexion at toe-off. In order to provide a stable 
mechanism over the widest range possible, knee linkage 
lengths were optimized through a computer program, (see 
Figure 1) 




Fig. 1: Conceptual paediatric running prosthesis. 



of rigid body segments were calculated by modelling 
human segments as frusta [14]. Constraint of head-arms- 
trunk segment rotations were imposed to counteract 
exaggerated hip abduction/adduction and internal/ 
external hip rotations. Joint rotation limits were based on 
clinical data [16]. (see Figure 2) 



3-OifnensJonol CoU Sifriulolton Model 




fte»loo/e«leo5km obduclion/odductlon 

plOOlOfflexIoo/dOrsiUeWoo volqul/vargl 

Fig. 2: Three-dimensional gait model (right). 

Three-dimensional positional data were acquired 
(VICON, 50Hz, low-pass filtering at 6Hz) from an able- 
bodied female running subject (9 year old), providing: i) 
initial position, ii) initial velocity, and hi) joint moments 
about all axes. Moment data and initial positions were 
used to drive the model simulating normal running gait. 
The model was then modified to approximate the 
prototype prosthesis and AKP gait by adding components 
and removing ankle and knee joint moments drivers. 

Evaluation 



Exaggerated knee flexion occurs during the swing phase 
of fast gait with conventional prostheses. A double- 
acting translational damper is proposed between the thigh 
and shank to provide swing phase knee flexion and 
extension control during walking and running gait, to 
enable faster shank return prior to heel contact, (see 
Figure 1) 

Most individuals with an AKA tend to extend their 
prosthetic knees through heel contact, creating 
uncomfortable axial loading on their residual limb, with 
little return of the energy absorbed. A proposed spring- 
damper device placed in the shank would absorb the large 
forces at heel-contact during running, and provide energy 
return at toe-off, improving gait symmetry, (see Figure 1) 



Simulation produced kinematic gait data to help in 
evaluating performance of the prototype prosthetic knee 
unit. Figure 3 demonstrates simulation results of the 
support and swing phases of running gait with the 
prototype prosthesis. 




AKA Gait Simulation with Prototype Prosthesis 
A three-dimensional model of the lower body was 
developed using DADS software [13]. The model, based 
on the model of Apkarian et al [14], incorporated four of 
the six determinants of gait: knee flexion, hip flexion, 
knee-ankle interaction, and lateral pelvic displacement 
[15]. The model consisted of both swing side and support 
side elements: foot, ankle, shank, knee, thigh, hip, and a 
common head-arms-trunk segment. Each hip and ankle 
joint was modelled as three orthogonal axes; each knee 
joint was modelled as a planar joint. Inertial parameters 



167 



Fig. 3: Simulations of prototype AKP gait during 

support phase (left) and swing phase (right). 

The prototype prosthesis exhibited stability under weight- 
bearing by limiting support-phase knee flexion (see 
Figure 4). Improved control of swing phase knee velocity 
and range during running was demonstrated (see Figure 
5) by the limited knee flexion range. The ability to 
achieve proper control of impact and support was 
demonstrated (see Figure 6) by the Imntation of COM 
displacement. 



RESNA’94 • June 17-22, 1994 



153 



GAIT SIMULATION and PROSTHESIS DESIGN 



RUNMNG SMULATION - SifK^-S^Dpcrt 




Fig. 4: Support phase knee flexion during running gait 

simulations. 



Summary 

A umque approach has been used to design and evaluate 
lower-limb prosthetic components. Quantitative gait 
aneilysis data have been combined with mechanical 
design methodology and computer simulation techniques 
to design and evaluate a new prototype pasdiatric above- 
knee endoskeletal running prosthesis. Results of both 
evaluation method and prosthesis are promising. 

Acknowledgments 

The author would like to thank The Hugh MacMillan 
Rehabilitation Centre and the Natural Science and 
Engineering Research Council of Canada for their 
flnancial support. 



RUNNING SMULATION - hilJol 47% of Swhg Phose 



References 




Fig. 5: Swing phase knee flexion during running gait 

simulations. 

RUNMNG SMULATION — Vertical COM Movement (sinqJe“St<3port) 




Fig. 6: Vertical COM movement during miming gait 

simulations. 

In comparing the running simulations between the 
prototype and the conventional prosthesis, the prototype 
demonstrated: 

1 . Slightly increased resistance to COM movement. 

2. Significant improvement in vertical COM move- 
moit. 

3. Decreased medio-lateral control. 

4. Increased knee stability during the support phase. 

5. Improved control of swing-phase knee flexion range. 

6. Modifiable vertical ground reaction force and 
vertical COM properties. 



RESNA’94 • 



[1] Bennett Wilson, A. Jr. (1989) limb Prosthetics (6th EcfilionsJ. Demos 
Publications; New York. 

[2] Beck, R.J., Andriacchi, T.P., Kuo, KN., Fcrmicr, R.W.. and Galantc, J.O. 
(1981) Changes in the gait patterns of growing children. J. Bone Joint Surg. 
63A, 1452-1456. 

[3] Norlin, R., Odcnrick. R, and Sandlund, B. (1981) Development of gait in 
the normal child. J. Fed Orthop. 1, 261-266. 

[4] Hoy, M.G., WhiUpg, W.C., and Zemicke, R.R (1982) Stride kinematics 
and knee joint kinetics of child amputee gait Arch. Phys. Med Rehab. 63, 
74-82. 

[5] Zemicke, R.F., Hoy, MG., and Whiting, W.C. (1985) Ground reaction 
forces and center of pressure patterns in the gait of children with amputation: 
preliminary report Arch. P^s. Med Rehab. 66, 736-741 . 

[6] Mensch, G., and Ellis, P.E. (1986) Running patterns of transfcmoral 
amputees: a clinical analysis. Pros. Orthot Int 10, 129-134. 

[7] Miller, D.I, Enoka, R.M., McCulloch. E.M, Burgess, E.M., Hutton, R.S., 
and Frankel, V.H. (1979) Biomedianical analysis of lower extremity aiTTputee 
extra-ambulatory activities. Final Technical Report to the Veterans 
Administration. Contract No. V5244P-1540AfA HOSP NY. 

[13] DiAngelo, D.J., Winter. D.A.. Ghista, D.N., and Ncwcombe, W.R. 
(1989) Performance assessment of the Terry Fox jogging prosthesis for 
above-knee amputees. J. Biomechanics 22, 543-558. 

[8] Mena, D., Mansour, J.M., and Simon. S.R. (1981) Analysis and synthesis 
of human swing 1^ motion duriiig gait and its clinical ^iplications. J. 
Biomeciumics 14, 823-832. 

[9] Pmdy, M.G., and Berme, N. (1988) 1. A numerical method for 
simulating the dynamics of human walking. 2. Synthesis of human walking: a 
planar model for single siQiport J. Biomechanics 2U 1043-1060. 

[10] Pandy, M.G., and Berme, N. (1989) Quantitative assessment of gait 
determinants during single stance via a three dimensional model-part 1. 
normal gait -part 2. pathological gait. J. Biomechanics 11^ 717-733. 

[11] Zamig)!, M.Y., and Radcliffe, C.W. (1976) Simulation of swing phase 
dynamics in above-knec prostheses. J. Biomechanics 9, 283-292. 

[12] Tsai, C.S., and Mansour, J.M. (1986) Swing phase simulation and 
design of above knee prostheses. J. Biomech. Eng. 108, 65-72. 

[13] Dynamic Analysis and Design of Systems', Conqiuter Aided Design 
Software Inc., P.O. Box 203, Oakdale. Iowa, 52319, USA 

[14] A|kanan, J., Naumann, S., and Cairns, B. (1989) A threesfimensional 
kinematic and dynamic model of the lower limb. J. Biomechanics 22, 143- 
155. 

[15] Saunders. J.B., Inman, V.T., and Ebcrhart, H.D. (1953) The major 
determinants in normal and pathological gait J. Bone Joint Surg. 35, 543- 
558. 

[16] Rothstein, J.M., Roy, S.R, and Wolf, S.L. (1991) The Rehabilitation 
Specialist's Handbook. F. A. Davis Company: RiiladelfMa. 



Alan R. Morris 



Gait Laboratory, Hugh MacMillan Rehabilitation Centre 
350 Rumsey Road, Toronto, ONT., CANADA 



M4G-1R8 

Tel: (416)425-6220 ext. 515 
Fax: (416)425-1634 



1.6S 



'1 \ 

June 17-22, 1994 



154 



UPPER EXTREMITY NET JOINT FORCES AND MOMENTS DURING WHEELCHAIR PROPULSION 



Rick N. Robertson Rory A. Cooper^’^, 
Biomechanics and Human Engineering Laboratories 
California State University 
Sacramento, CA 95819 



Greg J. Ensminger^ Ken J. Stewart^ 

Department of Rehabilitation Science & Technology 
School of Health and Rehabilitation Sciences 
University of Pittsburgh, Pittsburgh, PA 15261 



O 

ERIC 



The purpose of this study was to determine the net 
joint forces and moments at the wrist, elbow and 
shoulder during wheelchair propulsion and to 
show how these forces were related to joint 
position. Two experienced male wheelchair users 
pushed a Quickie 1 wheelchair secured to the 
CSUS wheelchair dynamometer Etted with the 3- 
chaimel SMART^*'®®^ on its right side. Video 
and force data were simultaneously collected for 
speeds from 1.34 to 1.79 m/s for 3 minutes for 
each subject Subsequent calculations of joint 
moments were completed by combining kinematic 
variables, push rim forces and anthropometric 
data. The data showed that the forces at each joint 
varied for both subjects in terms of peak forces, 
where they occurred in the propulsion phase and 
how quickly they developed. Also, the peak net 
joint moments occurred at different joint angles 
for both subjects from trial to trial. Considerably 
more data must be gathered and analyzed before 
definitive statements can be made about joint 
stresses and before this information can be utilized 
by the clinician. 

Introduction 

A determination of net joint forces and moments 
acting at a joint during locomotion provides the clinician 
and researcher with information related to the level of 
stress borne by the joint structures. The upper extremity 
during wheelchair propulsion has not been studied in this 
manner due to the lack of adequate instrumentation for 
determining push rim forces at the hand. The 
SMART'^*‘“> (1) allows push rim forces to be 
determined during wheelchair propulsion (2). Utilizing 
this information Cooper et. al. (3) presented a technique 
for determining net joint forces and moments utilizing an 
inverse dynamics approach. The purpose of this study 
was to determine the net joint forces and moments at the 
wrist, elbow and shoulder during wheelchair propulsion 
and to show how these forces were related to joint 
position. 

Background 

The kinematics of wheelchair propulsion have been 
studied by a number of researchers (4). Kinematic data 
by itself does not provide sufficient information for the 
clinician to implement appropriate rehabilitation 
intervention strategies or the engineer to incorporate this 
information into wheelchair design changes. A number of 
researchers have attempted to study joint forces during 
wheelchair propulsion but they have not been successful 
due to the lack of adequate instrumentation for ^ 
determining push rim forces at the hand. The 



SMART^^®®^ (1) is the first device which does this. The 
importance of evaluating net joint forces and moments is 
that it allows the clinician and researcher to study the 
level of stress experienced by the joint structures during 
propulsion. These forces can then be studied for different 
speeds of propulsion, injury level, user experience, and 
wheelchair type and fit. This is the first study known to 
the authors which presents realistic and accurate joint 
forces during wheelchair propulsion. The primary 
objective of this work is to provide appropriate 
information for the clinician to use in the rehabilitation 
process— designing training protocols, modifying stroke 
mechanics, and for the engineer to make appropriate 
changes to the wheelchair which allow the user to 
optimize efficiency and reduce trauma to the upper 
extremity. 

Methods 

Two experienced male wheelchair users pushed a 
Quickie 1 wheelchair which was secured to the CSUS 
wheelchair dynamometer (5) and fitted with the 
SMART^^®®^ on its right side at 1.34 to 1.79 m/s for 3 
minutes. Subjects monitored their speed by viewing a 
digital tachometer mounted at the front of the 
dynamometer. Data were collected at 75 Hz per channel 
for approximately 10 complete strokes near the end of the 
3 minutes. Processing of the signals from the beam- 
mounted strain gauges resulted in a determination of the 
forces in the X and y direction and moment about the z- 
axis (x*anterior-posterior, y-superior-inferior, z-medial- 
lateral). An optical encoder provided information about 
angular motion of the wheel. Simultaneous video data 
were collected at 60 Hz from a right sagittal view of each 
subject Video data were synchronized with the force 
data utilizing a reset of the optical encoder at top dead 
center. For analysis, the video data were interpolated to a 
time base of 75 Hz utilizing a spline function to match the 
force data. Subsequent calculations of joint moments 
were completed by combining kinematic variables, push 
rim forces and anthropometric data utilizing a method 
described by Coopca* et al. (3). Joint angle data were 
determined from ^gitized values using a Peak 5 System 
(Peak Performance Technologies, Inc.). Data were 
analyzed for 3 complete strokes for both subjects with 
routines written using MATLAB (Mathworks, Inc.). 

Results 

The push rim force data from this study has been 
presented elsewhere (2). Representative curves of forces 
and moment are given Figure 1. 

Peak Values . To evaluate the data, net forces and 
moments at the wrist, elbow and shoulder were analyzed 
in terms of maximum values (MaxFx (N). MaxFy (N)» 



RESNA ’94 • June 17-22, 1994 



155 



WHEELCHAIR JOINT MOMENTS 



ERIC 



MaxMz (N-m)), percent of propulsion stroke where 
maximum occurred (%strkFx, %strkFy, %strkMz), rate of 
rise of forces (RORFx (N/s), RORFy (N/s)), and impulse 
values— linear for Fx (ImpFx (N*s)) and Fy (LnpFy 
(N*s)) and angular for Mz (ImpMz (N-m*s)). Table 1 
presents the results from this analysis. 




niM(iec) 

Figure 1. Forces and moment at the Wrist, Elbow and 
Shoulder for Stroke 1 of Subject 1. 



The results of this part of the analysis showed that the 
maximum joint forces in the x<direction (maxFx) were 
similar for SI and S2 and increased going from the distal 
to proximal joint, (wrist to shoulder) SI developed 
considerably larger forces in the y-direction (maxFy) at 
aU joints. STs maxFy forces were larger than those in the 
X direction whereas S2 exhibited maximum Fy forces 
which were similar or smaller in value than the maximum 
Fx forces. The part of the propulsive phase where peak 
forces occurred were generally earlier for the Fx forces 
than the Fy forces for both subjects at all joints. Peak Fx 



RESNA’94 • 



forces occurred approximately 1/3 of the way into the 
propulsive phase whereas peak Fy occurred between 50> 
60% into this phase. The ROR of both forces varied 
considerably between subjects and from joint-to-joint S2 
developed greater moments than SI at all joints. S2 
produced the largest moments at the elbow whereas for 
SI the wrist moment was the greatest. 



Table 1. Net Joint Force and Moment Variables 
Averaged over 6 strokes for both subjects. 



Wrist 


MaxFx 


%strkFx 


RORFx 


ImpFx 




31.9 


36.4 


123.9 


13.3 




C3.7) 


(11.1) 


(38.6) 


(1.6) 




MaxFy 


%strkFy 


RORFy 


ImpFy 




-48.9 


53.2 


-126.5 


22.9 




(21.6) 


(15.0) 


(54.5) 


(8.5) 




MaxMz 


%strkMz 




ImpMz 




26.4 


62.5 




13.3 




(7.6) 


C4.3) 




(6.3) 


Elbow 


MaxFx 


%strkFx 


RORFx 


ImpFx 




37.5 


36.5 


144.7 


14.0 




(3.1) 


(10.8) 


(48.1) 


(1.7) 




MaxFy 


%strkFy 


RORFy 


ImpFy 




-47.9 


56.3 


-113.9 


18.6 




(19.5) 


(4.0) 


(51.4) 


(8.6) 




MaxMz 


%strkMz 




ImpMz 




25.4 


67.1 




11.9 




(14.8) 


(11.3) 




(9.0) 


Shoulder 


MaxFx 


%strkFx 


RORFx 


ImpFx 




46.3 


30.6 


234.2 


15.8 




(4.2) 


(12.4) 


(118.5) 


(2.0) 




MaxFy 


%strkFy 


RORFy 


ImpFy 




-443 


55.8 


-110.5 


15.0 




(21.1) 


(15.4) 


(59.1) 


(7.9) 




MaxMz 


%strkMz 




ImpMz 




20.1 


44.0 




9.3 




(151) 


(26.6) 




jm 



Joint Moment-Angle Relationships . The angles were 
all determined in the sagittal plane and represented 
flexion and extension at the 3 joints. The convention for 
joint angles was that 180 degrees at the elbow represented 
full extension while at the wrist this was with the hand in 
the neutral position (flexion less than 180 degrees and 
extension greater than 180 degrees). Joint angles at the 
shoulder were determined between the arm and the trunk, 
with zero at the point where the trunk and arm were 
aligned. Less than zero indicated that the arm was 
anterior to the trunk and greater than zero was the arm 
posterior to the trunk. Figure 2 shows typical moment- 
angle plots for all 3 joints. 

Both subjects showed some variability from stroke to 
stroke in terms of how the moment was produced 
throughout the range of motion, there were considerable 
di^erences between the subj'TOts. The maximum net 
shoulder moment for both subjects occurred between 20- 
40^ of extension. Both Subjects showed a rapid rise in 

79 

. 1 

June 17-22, 1994 



156 



WHEELCHAIR JOINT MOMENTS 



elbow extensor moment at the beginning of the stroke, 
with the elbow at about 120^. This moment value began 
to decrease at approximately 150^ for SI and 170^ for S2. 
At the wrist, the peak moment occurred at close to 220^ 
for SI and close to 190^ for S2. 




ShmhfarAp|>p(Pegwc») 



the elbow for S2 indicate that these joint structures were 
being stressed to a greater extent than at the other joints. 
At the wrist, peak moments occurred when the hand was 
close to its most stable position for S2 (i.e., close to 180^) 
and in a slightly extended position for SI. The results 
from this study show that net joint moments and forces 
are considerably variable within a subject and between 
subjects. It is therefore in^rtant to deteimine how 
different conditions of propulsion, such as speed, level of 
injury, user experience, fit in the chair, and type of chair, 
affect these joint forces. 

Summary 

Considerably more data must be gathered and 
analyzed before definitive statements can be made about 
joint stresses and before this information can be utilized 
by the clinician. More work needs to be done to 
accurately model the center of pressure on the hand as 
this will influence the net joint moment at the wrist. 
More subjects must be tested under varying conditions 
and the analysis needs to be extended to the other planes 
of motion. Similarly, changes in wheelchair design can 
only be implemented once we understand how various 
design features, conditions of propulsion and an 
individual's physical characteristics influence net joint 
forces and moments. 

References 

1. Asato, K.T., Cooper R-A, Robertson R.N., & Ster, JJF., (1993). 
SMART^^®®* Development and testing of a system for measuring 
manual wheelchair propulsion dynamics. IEEE Transactions on 
Biomedical Engineering, Vol. 40, No. 12. 

2. Robertson, R.N., & Cooper, R. A, (1 993). Kinetic 

characteristics of wheelchair propulsion utilizing the 
SMART'^^®®*. Proceedings of the 17th Annual Meeting of the 
American Society of Biomechanics, Iowa City, Iowa, pp. 202-203. 

3. Cooper R.A, Robertson, R.N., & VanSickle DP., (1993), A 
recursive back propagation algorithm for computing net muscle 
mcmients and net joint fortes. Proceedings 16th Annual RESNA 
Conference, Las Vegas, NV, pp. 277-279. 

4. Sanderson D.J. & Sommer m, H.J., (1985). Kinematic features 
of vriieelchair propulsicxi. Journal of Biomechanics, 18(6), 423* 
429. 

5. Cooper R.A., (1989) Simulating wheelchair racing. Proceedings 
of the 12th Atmual RESNA, New Orleans, pp.450-45 1 . 



Figure 2. Joint Moment- Joint Angle Graphs of the Wrist, 
Elbow and Shoulder for Stroke 1 of Subject 1. 

Discussion 

The data showed that the forces at each joint varied 
for both subjects in terms of peak forces, where they 
occurred in the propulsion phase and how quickly the 
forces developed. Also, the peak net joint moments 
occurred at different joint angles for both subjects and the 
way in which the joint moments were produced 
throughout the full range of motion was different from 
trial-to-trial and between subjects. The larger moments at 



Acknowledgments 

Partial funding for this research was provided by the U.S. 
Department of Veterans Affairs Rehabilitation Research 
and Development Services (Plroject#B686-RA) through 
the Edward Hines Jr. D.V.A. Hospital and the U.S. 
Department of Education, Rehabilitation Services 
Administration (H129E00005). 

Rick N. Robertson Biomechanics Laboratory 
Mailstop 6073 

California State University, Sacramento 
, ^Sacramento, CA 95819-6073 



OPY AVAILABLE 



171 

RESNA ’94 • June 17-22, 1994 



157 



DIFFERENTIAL PRESSURE WALKING ASSIST 



Douglas F. Schwandt^, Robert T. Whalen’*’*, Gregory A. Breit** and Charles G. Burgar* 
‘*‘DVA Rehabilitation Research and Development Center, Palo Alto, California 
’*”*'Life Science Division, NASA Ames Research Center, Moffett Field, California 



ABSTRACT 

We have applied the concept of differential pressure to 
reduce the loads on the lower body during upright 
Standing and walking. The first prototype of a 
differential pressure walking assist device has been 
constructed to test the feasibility of the concept and to 
investigate cardiovascular effects on able-bodied 
subjects during standing. The device consists of an 
inflatable conical reinforced vinyl skirt attached to a 
circular base holding a force plate. Buoyant forces 
between 0 and 100% of body weight were 
demonstrated without adverse cardiovascular effects in 
this study population. 

STATEMENT OF THE PROBLEM 

An improved walking assist device, which 
comfortably reduces the ground reaction force and 
lower limb muscle and joint forces, is needed for 
patients recovering from neurologic diseases and 
orthopedic procedures. 

BACKGROUND 



RATIONALE 

We have applied the concept of differential pressure to 
reduce the loads on the lower body during upright 
standing and walking. To create the pressure 
differential, the subject's lower body is enclosed in an 
inflatable chamber isolated from the upper body at the 
waist by a flexible, air-tight seal. 

During quiet standing tests, the ground reaction force, 
the vertical force between the ground and the plantar 
surface of the foot for an upright subject, decreased 
proportionally with increasing pressure. The ground 
reaction force is the vector sum of the downward force 
due to gravitational acceleration and the upward 
differential pressure force. The force due to air 
pressure is equal to the product of the pressure 
difference and the waist cross-sectional area. The line 
of action of the resultant force passes through the area 
centroid of the waist cross-section, which is near the 
body center of mass. An additional upward shear 
force is created by the waist seal against the subject 
due to area of the flexible waist seal exposed to the 
pressure. 



A variety of techniques have been used to help 
partially suspend a patient's weight during walking 
rehabilitation, including water immersion, parallel 
bars, walkers, overhead suspensions, and therapist- 
supported waist belts. Most of these techniques have 
disadvantages associated with them, in addition to 
interfering with normal gait: 

• v/ater immersion - inconvenience, possible 
infection of open wounds or incisions, viscous 
drag, and difficulty in modulating forces 

• parallel bars and walker - strength required of the 
patient's upper body 

• hand-held waist belts - physically stressful for the 
therapist supporting a patient's weight; possible 
patient falls 

• overhead suspension harness - locally high interface 
contact pressures. 

Lower body differential pressure chambers large 
enough to enclose a treadmill have been developed at 
the Life Science Division of NASA- Ames Research 
Center [1,2,3]. These chambers are being used in 
research to investigate the qipplication of differential 
pressure to exercise in space, and to simulate hypo- 
and hyper-gravity locomotion [4]. 



Since the upper and lower body air pressures are 
uniformly distributed over the body surfaces, the 
resultant force is not felt as a localized force lifting up 
on the body. The subject simply feels lighter, 
similar to being buoyed up by waist-deep water. The 
lifting force can easily be controlled by adjusting 
chamber pressure to adapt to an individual patient's 
needs as he or she progresses through rehabilitation. 

DESIGN 

The first prototype of a differential pressure walking 
assist device has been constructed to test the 
feasibility of the concept and to investigate 
cardiovascular effects on able-bodied subjects during 
standing. The device consists of an inflatable conical 
reinforced vinyl skirt attached to a circular plywood 
base holding a force plate (see Figure 1). The subject 
steps through the waist seal onto the force plate, 
pulls the skirt up, and positions the elastic neoprene 
flap seal at the waist before inflation. 

A computer controls a servo-valve connected to a 
blower which regulates the pressure in the chamber, 
while simultaneously recording and displaying 
pressure and ground reaction force. 







RESNA’94 • June 17-22, 1994 



158 



Differential Pressure Walking Assist 




DEVELOPMENT 

Prototype development is continuing to allow 
enclosure of a treadmill at the bottom of the Inflatable 
conical skirt. With assistance from a therapist, a 
patient in a wheelchair will be able to roll up a ramp, 
enter the skirt through an air-tight zippered entrance, 
and stand up on the treadmill with the aid of external 
handrails. The skirt will be inflated to the desired 
pressure before starting the treadmill. 

If results of ongoing work continue to be 
encouraging, a commercial version of the device will 
be developed In collaboration with an industrial 
partner. 

DISCUSSION 

The use of lower body positive pressure is a 
comfortable way to reduce the effects of gravitational 
forces. Since the upward resultant air pressure force 
acts at or near the center of mass of the body, we 
anticipate walking in the device will resemble normal 
gait but with proportionally reduced musculoskeletal 
forces. There Is contact with the patient only at the 
waist seal. The arms are free to move as pressure 
and shear force in the seal do all of the lifting, and the 
waist seal provides front, back and lateral support to 
prevent falling. 



ERIC 



EVALUATION 

A pilot study of eight healthy consenting male 
subjects between the ages of 29 to 52 established the 
relationship between ground reaction force and 
pressure, and evaluated the cardiovascular responses. 
Blood pressure and heart rate were measured every two 
minutes during two protocols and the results 
compared to a supine resting baseline. In the first 
protocol the pressure was changed stepwise from 0 to 
45 mmHg and back to zero in 15 mmHg steps at two 
minute intervals. In the second protocol, each 
subject's response to rapid changes in pressure was 
evaluated by pressurization to 45 mmHg for 15 
minutes and then rapid return to ambient pressure. 

Changes in heart rate and blood pressure were within 
the range of values observed between baseline and 
standing. 

At 45 mmHg, mean reduction in vertical ground 
reaction force was 75%, and ranged from 60% to 
100%. Variation in per cent reduction of ground 
reaction force is likely due to variability in body 
habitus and waist seal placement on the body. We 
also found as expected that the pressures necessary to 
substantially unload subjects were significantly lower 
than pressures predicted by waist cross-sectional area 
alone, due to the upward shear force from the waist 
seal. 



The first trials of this device have been limited to 
able-bodied healthy adult subjects. Levels of lower 
body positive pressure necessary to unload subjects 
legs are relatively benign. For instance, the 
maximum pressure of 45 mmHg attained in this 
initial study is equivalent to the hydrostatic pressure 
of two feet of water. 

Additional studies are planned to further determine the 
efficacy and safety of the device for use in 
rehabilitation before testing with patients. 

REFERENCES 

1 . Whalen, R.T.; Hargens, A.R.; Schwandt, D.F.; 
Watenpaugh, D.E. (1991) Musculoskeletal 
loading or unloading with differential lower body 
pressure. Transactions of the 37th Annual 
Meeting of the Ortbop. Res. Soc. (ORS), p.628, 
Anaheim, March 4-7. 

2. Hargens, A.R.; Whalen, R.T.; Watenpaugh, 
D.E.; Schwandt, D.F.: Krock, L. (1991) Lower 
Body Negative ^essure to Provide Load Bearing 
in Space, Aviat, Space Environ, Med. 62:934-7. 

3 . United States Patent # 5, 133,339: Exercise 
Method and Apparatus Utilizing Differential 
Pressure, Inventors: Robert Whalen and Alan 
Hargens. 

4. Whalen, R T,; Breit, GA.; Schwandt, D.F. (in 
preparation). 



173 

RESNA’94 • June 17-22, 1994 



159 



Differential Pressure Walking Assist 

ACKNOWLEDGMENTS 

The initial development of the differential pressure 
walking assist is a collaborative effort between the 
Life Science Division at the NASA Ames Research 
Center (NASA-ARC) and the Human Machine 
Systems Group at the Palo Alto DVA Rehabilitation 
Research and Development (RRD) Center. 

Fabrication of machined parts, the wooden base, and 
pressure system connections are the work of James 
Anderson, modelmaker machinist at the RRD Center. 
Jeff Emery and Josh Beach contributed to the 
development of the pressure control and data 
acquisition systems at NASA-ARC. The neoprene 
waist seal was constructed by Stan Stugen at Stan's 
Skin Diving Shop in San Jose, California. 



Douglas F. Schwandt, MS 
Biomedical Engineer 

Rehabilitation Research & Development Center 
Department of Veterans Affairs Medical Center 
3801 Miranda Avenue (MS 153) 

Palo Alto, California 94304-1200 USA 
415/493-5000x4473 FAX: 415/493-4919 
INTERNET: schwandt@roses.stanford.edu 



O 



ERIC 160 




RESNA’94 • June 17-22, 1994 



Measuring Leg Motion Changes Following Vertical Vestibular Stimulation: 

A Case Study 

Katherine T. Samworth PT, BEE, James W. Fee, Jr. MS 

Applied Science and Engineering Laboratories (ASEL) 

A.I.duPont Institute, Wilimington, DE 19803 



ABSTRACT 

This laboratory is investigating the effects of vertical 
vestibular stimulation on lower limb motion in indi- 
viduals with cerebral palsy. A case study of an 18 year 
old subject with spastic diplegic cerebral palsy is pre- 
sented. The subject’s active and passive leg motion 
was measured before and after vestibular stimulation. 
Seventeen attributes were calculated from these mea- 
surements. Statistical analysis showed a significant 
difference (p<.05) in the measurements before and 
after stimulation in 16 out of the 17 attributes. The 
results indicate that this type of vestibular stimulation 
may have had an effect on this subject’s active and 
passive leg motion. Further implications are noted. 



INTRODUCTION 

When designing a therapy program for individuals 
with neurological impairments, therapists will often 
include some type of vestibular stimulation in their 
treatment regime. Researchers claim that vestibular 
stimulation offers benefits such as normalization of 
muscle tone, improved postural alignment and 
improved balance reactions[2][3][6]. Stimulation is 
often achieved by placing the individual in a swing or 
on a platform and spinning them. In an alternative 
method, hippo therapy, the individual is stimulated by 
being moved primarily in the vertical direction. In a 
review of hippo therapy literature, Engel[l] reports 
effects such as decreased spasticity, improved trunk 
control and increased walking stability. 

OBJECTIVE 

The authors and their co-workers have built a platform 
that simulates the vertical movements of a trotting 
horse in order to study it’s effect on individuals with 
cerebral palsy. The subject ‘rides’ on the vertically 
moving platform and is then measured to detect 
changes in active or passive leg motion. Use of this 
platform allows the direction, frequency and magni- 
tude of stimulation to be controlled for more objective 
evaluation of any effects that it may produce. 

METHOD 

The subject to be presented in this case study is an 18 
year old female with spastic diplegic cerebral palsy. 
She attends high school and is essentially independent 
in her activities of daily living. She walks with some 
gait deviations and has spasticity in her quadriceps. 

The vertical motion platform used in this case study is 



shown in Figure 1. The subject was placed on the plat- 
form in a secured chair. The platform was then oscil- 
lated vertically with an amplitude of 3.5 inches and a 
frequency of 1.57 Hertz. 




The leg position information was collected using a 
magnetic sensing device, the 3SPACE® Isotrak® by 
Polhemus Navigational Sciences Division, McDon- 
nell Douglas Electronics Company. The device uses 
the principle of low-frequency magnetic field technol- 
ogy to determine position and orientation of a sensor 
relative to a source. The subject was seated with the 
upper leg held fixed. The sensor was then placed at the 
lateral malleolus in order to record knee angle. 

The subject was asked to perform three leg motion 
tests before and after fifteen minutes of stimulation. 
First, the leg drop pendulum test was performed to test 
passive motion [4]. The leg was held so that the knee 
was extended as far as the subject could comfortably 
tolerate. When the subject’s quadriceps muscle 
appeared to be relaxed, the leg was then dropped. 
Position information was recorded until the leg 
stopped moving. Next, the ‘kick’ test was performed. 
The subject was asked to kick a hanging ball as hard 
as she could. Lastly, the ‘swing’ test measured the 
motion as the subject was asked to swing her leg back 
and forth for ten seconds. Each test was repeated six 
times before and six times after stimulation. 

DATA ANALYSIS 

The data analysis included looking at seventeen mea- 
sures of performance based on information from the 
three tests. The attributes measured are as follows: 



175 

RESNA’94 • June 17-22, 1994 



161 



Leg notion changes following vestibular stin. 

1 . For the pendulum test: 
undamped natural nrequencyCsec'^); 7 

P 

{ip S3 time of peak position) 
damping ratio; 

(jc 3s tan^ (d^i^ j and = the rise time) 

settling tlme(sec): t, =3 the time for the leg to settle 
within 1% of 0^; (0^ a final resting angle) 
maximum angle(degrees): 0^ => 0^ ~ 0^ 

( 0^ initial angle, 0^, a final resting angle) 
maximum drop(degrees): 9^ ® 

( 0p3s angle at first pe^) 

averse velQcity(radians/second): Og ^ slope of the 
first order fit of <0^ (0 ^ 9 ^ (0 > 
maximum velocity(radians/second): s point 

where J^<e.(o -0^(,)) ^ q 

maximum spectral density; p was calculated by 
^plying a fast fourier transform on the oscilla- 
tory part of the signal. This was done by remov- 
ing the DC component, multiplying by a 
hamming window, filling to 4096 points with 
zeros, and dividing each value by the number of 
points in that particular trial. From this the 
maximum energy is found. 

2 . Kick test attributes: 

Kick is defined as a ballistic movement: continuous 
and in one direction, 
maximum range of motion(degrees): 

MROM = 0.-0 
' P 

( e. = initial angle and 0^ = peak angle) 

maximum deflection(degrees): d =0-0 

/W 0 c 

( 0^ = the angle where the velocity = 15 ra^sec 

and 0^ = the final position of dte kicking 
stroke) 

average velocity(radians/second): v* - slope of the 
first order fit of <0.(0 =0 (0) s 0 

maximum veiocItyCradians/second): s point 

.2 

where ^<0^(o.0^(o) ^ 0 

3 , Swing Tfest AtUibutes: 

maximum range of motion(degrees): MROM b 
difference between the max, and min. angle 
maximum spectral density: p (same description as 
for the pendulum test) 

total power: P = the integral of the spectral density 
curve 




frequency at maximum power; / 
moment of inertia: M = Area moment of inertia of 
the power spectral density curve about the fre- 
quency of maximum power. This moment rep- 
resents a measure of the dispersion of 
frequencies. 

RESULTS 

Visual inspection of the data (see Figures 2 , 3 and 4) 
indicates a change in passive and active leg motion 
following stimulation. Mean values and standard 
deviations for particular attributes are presented in 
tables 1, 2 and 3 . Statistical methods were based on 
those outlined in Ottenbacher’s chapter “Statistical 
Analysis of Single System Data”[ 5 ]. Autocorrelation 
calculations were performed on the data in order to 




Figure 2 . Pendulum Test 

nile out serial dependency (due to practicing or learn- 
ing effects). Since no serial dependency was detected, 
w analysis of variance (f-test) was used to test for sta- 
tistical significance. A significant difference was seen 
in thirteen of the attributes at the 99%confldence 
level. Three attributes were significant to the 95 % 
confidence level and one was significant to the 90 % 
confidence level. Ibbles 1, 2 and 3 give information 
regarding the mean (p), the standard deviation (o) and 
the p-values for each of the attributes. 





Figures. Kick Test 



RE$NA’94 • June 17-22, 1994 



162 



Leg motion changes following vestibular stim. 



b«(or* ttkmiUUon 

lOr « 1 » 




100 200 300 400 500 «0O 700 



sUmulaUon 




Table 1: Pendulum test 



Attribute 


labefore 


abefore 


\x after 


G after 


p 


Max angle(0;„) 


64.30 


2.61 


70.75 


2.57 


.01 


Max Drop (0^) 


51.11 


5.16 


92.73 


9.30 


.01 


Ave Vel (v^) 


3.65 


0.26 


4.94 


0.32 


.01 


Max Vel 


8.33 


1.04 


10.34 


0.41 


.01 


Damping (^) 


2.0 


0.00 


0.72 


0.11 


.01 


Nat Freq (coj ) 


11.05 


1.08 


8.27 


0.8- 


.01 


Sett time (t^) 


3.18 


0.31 


7.36 


0.67 


.01 


Max Dens (p) 


0.29 


0.46 


1.08 


0.01 


.01 



Table 2: Kick test 



Attribute 


xbefore 


abefore 


\i after 


G after 


p 


Max ROM 


28.52 


1.01 


31.10 


1.35 


<.01 


MaxDefl (d,^) 


32.55 


3.38 


29.04 


1.57 


<.05 


Ave Vel (v^) 


2.09 


0.04 


2.31 


0.12 


<.01 


Max Vel (v„) 


3.99 


0.33 ^ 


4.27 ^ 


0.24 


<.10 



Table 3; Swing test 



Attribute 


ibefore 


abefore 


|i. after 


G after 


p 


Max ROM 


22.00 


2.45 


41.07 


4.92 


<.01 


Max Spec (p) 


0.79 


0.23 


2.80 


0.60 


<.01 


Total Pow (P) 


0.12 


0.03 


0.42 


0.11 


<.01 


Freq MP (/) 


1.23 


0.06 


1.30 


0.03 


<.05 


Mom Inert (M) 


0.26 


0.01^ 


0.29 


0.02 


<.05 



It is interesting to note that, before stimulation, the 
subject was asked to swing her right leg. She replied 
“I can’t unless 1 swing both legs together”. She 
appeared unable to dissociate the movement of one 
leg from the other. After stimulation, when asked 
again, she immediately began swinging her right leg 
only. 



DISCUSSION 

Results indicate that there appears to be a difference in 
this subject’s leg motion before as compared to after 
stimulation. Some of the effects seen following stimu- 
lation, such as decreased damping of motion, 
increased active and passive range of motion and 
increased velocity of motion indicate a possible 
decrease in muscle stiffness. This may be related to a 
decrease in muscle spasticity. In addition, the “catch” 
that is seen before stimulation in the pendulum test 
(see Figure 2.) as the leg begins to fall disappears after 
stimulation. This could indicate a decrease in spastic- 
ity as well. Further research with larger sample popu- 
lations is crucial in order to make decisions about the 
efficacy of treatment regimes that include vestibular 
stimulation. 

REFERENCES 

1 Engel, B.T., ‘The Horse as a Modality for Occupa- 
tional Therapy” Am. J. OccuP. Ther. 38:4 1 -47, 1984. 

2 Farber, S. D., Neurorehabilitation: A Multisensorv 
Annroach . Philadelphia, Pa.: W. B Saunders Co., 
1982:132-144. 

3 Fiebert, I.M., and Brown, E., “Vestibular Stimula- 
tion to Improve Ambulation after Cerebral Vascular 
Arcident”. Phvs. Ther . 1979; 59(4):423-426. 

4 Katz, R.T., Rovai, G.P., Brait, C„ Rymer, W.Z., 
“Objective Quantification of Spastic Hypertonia”, 
Arch. Phvs. Med. Rehab. . 1992, 73: 339-347. 

5 Ottenbacher, K. J. Evaluating Clinical Chanite: Strata 
p.pjp.s for Occiinational and Physical Therapists . Will- 
iams & Wilkins, Baltimore, 1986, p. 178. 

6 Trombly, C.A., Occupational Therapy for Physical 

nvsfiinction . Baltimore: Williams & Wilkins, 

1984:65-66. 

ACKNOWLEDGEMENTS 

Funding for this research has been provided by 
Nemours Research Programs. 

ADDRESS 

Katherine T. Samworth 

Applied Science and Engineering Laboratories 
A.LduPont Institute, 

PO Box 269, Wilmington., Delaware 19899 

email:samworth@ asel.udel.edu 
phone:302-65 1-6830 



177 



ST COPY AVAILABLE 



RESNA >94 • June 17-22, 1994 



163 



FUNCTIONAL ASSESSMENT OF A TRICEPS ORTHOSIS FOR C5/C6 TETRAFLEGU 



Allen W. Wiegner and M. Margaret Wierzbicka 
Spinal Cord Injury Service, Brockton/West Roxbury VA Medical Center 
and Department of Neurology, Harvard Medical School 
Boston, MA 02132 



ABSTRACT 

Persons with spinal cord injury at the C5/C6 
cervical level typically have relatively well pre- 
served biceps function, but minimfll or no 
voluntary control of triceps. Our previous work 
has showed that this results in deficiencies in speed 
and accuracy of elbow movem^ts, as well as a 
reduction in the reachable woikspace, and that 
these deficiencies can be corrected by the addition 
of constant extensor torque and damping at the 
elbow. In this rqjort we describe a prototype 
mechanical orthosis, worn at the elbow, and 
illustrate inqprovements in function obtained by 
users with CS/C6 tetraplegia. 



"artificial triceps," however, patients' movement 
speeds increased and their accuracy was 
indistinguishable from that of control subjects (2). 
We also used the torque motor to electronically 
emulate the addition of up to 1.0 Nms/rad of 
damping and the substitution of a linear extensor 
spring for the constant extensor torque. We found 
that the addition of damping {q)pears to contribute 
to accuracy and the subjective feeling of being in 
control; the linear spring increased oscillations and 
gave subjects a feeling of instability (3). The 
realization of this concept in an orthotic device 
requires a constant-torque spring to provide the 
ext^isor torque and the optional addition of a 
damping element. 



BACKGROUND 

Spinal cord injury (SCI) results in the loss of vo- 
luntary control of muscles and causes decreased 
mobility to an extent depending on the level and 
conq)leteness of injury. Persons with injuries at the 
C5/C6 cervical level, whose bic^s function is 
relatively well preserved, have little or no 
voluntary control of triceps. Without use of the 
triceps, which is the primary elbow ext^isor, they 
lose the ability to, for exanq>le, reach overhead. In 
addition, they face limits on the range, speed, and 
accuracy of arm movements; this loss of control of 
the arm occurs even in a task in which the biceps 
does the work. 

Because even a small amount of bicq>s spasticity 
can result in a chronic torque imbalance at the 
elbow joint with a resultant flexed, supinated 
forearm posture, a number of orthotic devices have 
been developed to reverse elbow flexion postures or 
contractures (see (1) for references). On the other 
hand, the control deficit during biceps tasks has 
received little att^tion. Our recent studies of fast 
elbow flexion movements in C5/C6 tetraplegics 
have revealed deficiencies in movement speed and 
accuracy compared with control subjects (2). 
Patients' movement times were, on average, twice 
those of control subjects, and their errors were 
twice as large when they attempted 10° 
movements. When we used a torque motor to 
provide a 2.5 Nm constant extensor torque as an 



RESNA ’94 



ORTHOSIS DESIGN 

The prototype orthosis is illustrated in Fig. 1. 
The four cuffs were fabricated of carbon fiber 
lamination braid and lined with T-foamT^. As we 
were unable to locate a commercially available 
rotational constant torque spring, we used a 
prestressed power spring which could be prewound 
to an appropriate torque (1.25 Nm each side) and 
maintained that torque within 10% over the limited 
(160°) range of usage. Damping was provided by 
adjustable linear air dampers (Airpot Corp.) 

ORTHOSIS EVALUATION 

Subjects were outpatients of the Spinal Cord 
Injury Service, selected on the basis of relatively 
normal biceps strength and little or no voluntary 
strength in triceps. 

Control test. To demonstrate improvement in 
control aspects of arm motion, one subject 
performed single-joint movements to a visual 
target, with and without the triceps orthosis (10 
trials under each condition). He sat in his own 
wheelchair with the arm supported on a table at 
shoulder height and strapped to a manipulandum, 
allowing horizontal rotation of the arm at the 
elbow. An oscilloscope displayed two traces: one 
with the initial and then final target level, and the 
other showing the subject's current arm position. 
The subject aligned with the initial target line then 
moved "as quickly and accurately as possible" to 




June 17>22, 1994 



164 



ASSESSMENT OF TRICEPS ORTHOSIS 



the final position when the target shifted. 
Movements of 10® and 30® were performed. 

Average (±SD) movement times decreased from 
312±34 msec to 241 ±31 msec in 10® movements 
and 468 ±49 msec to 294 ±35 msec in 30® 
movements. Acceleration and deceleration times 
were both improved. Movement accuracy was 
similar with and without the orthosis, although 
subjects moved much faster with the brace. These 
faster movements, made without the brace, would 
have resulted in large overshoots of the targets. 
These improvements in kinematic performance are 
similar to those of our previous studies (2,3). 

Reaching test. Three subjects performed reaches 
to a 60 cm high, 120 cm wide "target board", 
placed on a table 30 cm in front of their chest. 
Targets were 4 cm diameter holes arranged in 4 
rows of 7 columns, spaced 18 cm apart. A 
lightweight pen was strapped to the subject's index 
finger. He was instructed to bring his hand from 
an initial resting position on the table to each target 
in turn. Success was based on ability to place the 
pen into each hole and maintain that position. 

Results differed in each subject because this test 
of three-dimensional reaching requires coordinated 
movement and strength of several muscle groups, 
including shoulder muscles. Subject 1, with no 
triceps strength and fairly weak deltoid muscles, 
showed marked improvement in his ability to reach 
the higher targets while wearing the orthosis (Fig. 
2). Subject 2 had a trace of triceps but was limited 



by a biceps contracture in the tested arm; 
nevertheless he showed the ability to reach 
additional targets while wearing the orthosis. 
Subject 3 had a weak triceps and strong deltoids, 
and could reach all targets even without wearing 
the orthosis. This subject might not be a suitable 
candidate for the orthosis because he would show 
less functional improvement from its use. Overall, 
however, the orthosis brought more targets within 
reach, allowed faster reaches, and abolished arm 
collapse during overhead reaches. 

Side reaching test: Subjects were asked to 

extend the arm to the side, then abduct it as high as 
possible while seated in the wheelchair. Subject 1 
increased his vertical side reach from 129 to 142 
cm; subject 2 from 112 to 127 cm; and subject 3 
again showed no improvement. 

Propelling the wheelchair: Without the orthosis, 
subject 1 could extend his arm only to a gravity- 
assisted 48® before gripping the rim. With the 
orthosis, he could further extend to 23®, allowing a 
longer power stroke. 

Movement mimicking feeding: Subject I's task 
was to r^jetitively pick up a light object from the 
table (using tenodesis grip) and bring it to his 
mouth, then return it to the table. Wearing the 
orthosis did not result in an increase in speed over 
20 repetitions, but did result in fewer collisions 
with the chin, lips, and nose. 




FIG. 1. A prototype version of the triceps orthosis. Note the springs (in cylindrical cases) and the dampers. 

. '179 

RESNA ’94 • June 17-22, 1994 165 




ASSESSMENT OF TRICEPS ORTHOSIS 
DISCUSSION 

Basic research in neuromuscular motor control 
has recently led to recognition of deficits in control 
of arm movements of C5 tetraplegic subjects even 
in the absence of biceps spasticity or flexion 
contractures. This suggests that the applicability of 
elbow extension orthoses to the C5 tetraplegic 
population may be far wider than previously 
appreciated. 

In rehabilitation practice, particularly involving 
prosthetic or orthotic devices, the simplest solution 
is often the best. The overall goal of this research 
project is to develop a lightweight triceps orthosis, 
composed of passive mechanical elements, to 
in^rove control of the arm, allow faster 
movements with improved accuracy, and provide 
increased reaching capacity for persons with C5/C6 
tetraplegia. Alternative therapies to restore triceps 
function include tendon transfer surgery (4) and 
ftmctional electrical stimulation (5). 



o 


o 


o 


• 


• 


o 


o 


o 


• 


• 


• 


• 


• 


o 


o 


• 


• 


• 


• 


• 


o 


• 


• 


• 


• 


• 


• 


o 


(A) 














o 


• 


• 


• 


• 


• 


o 


• 


• 


• 


• 


• 


• 


o 


• 


• 


• 


• 


• 


• 


o 


• 


• 


• 


• 


• 


• 


o 



(B) 



Although the prototype orthosis meets these 
goals for improved function, substantial further 
development is required to minimize its bulk and 
maximize its comfort and convenience. Even more 
than splints developed for tenqjorary clinical 
purposes, an orthotic device intended for chronic 
use at home must excel in terms of convenience, 
long-term comfort, and attractiveness in order to be 
accepted and used. Current development efforts are 
directed toward eliminating the medial spring 
(which can get in the way during wheelchair 
propulsion), providing a pronation function, 
reducing the size and weight of the side bars, 
setting up a way to disable the spring mechanism 
during periods when the arm is inactive, and 
improving the donning/doffing procedure. 

REFERENCES 

1. Wiegner AW. Can basic science help improve 
arm function in C5 tetraplegia? J Am Paraplegia 
Soc 1993; 16: 75. 

2. Wierzbicka MM, Wiegner AW. Effects of 
weak antagomst on fast elbow flexion movements 
in man. Exp Brain Res 1992; 91: 509-519. 

3. Wiegner AW, Wierzbicka MM. Mechanical 
compensation for weak triceps in C5/C6 
tetraplegia. IEEE Trans Rehab Eng 1993; 1: 72-78. 

4. Freehafer AA, Kelly CM, Peckham PH. 
Tendon transfer for the restoration of upper limb 
function after a cervical spinal cord injury. J Hand 
Surg 1984; 9A: 887-893. 

5. Miller U, Peckham PH, Keith MW. Elbow 
extension in the C5 quadriplegic using ftmctional 
electrical stimulation. IEEE Trans Biomed Eng 
1989; 36: 771-780. 

ACKNOWLEDGMENTS 

This woik was supported in part by the 
Paralyzed Veterans of America Spinal Coid 
Research Foundation, Grant 146. 

Allen W. Wiegner, Ph.D. 

Spinal Cord Injury Service (128) 

VA Medical Center 
1400 VFW Parkway 
Boston, MA 02132 



FIG. 2. Subject 1 could reach more targets (filled 
circles) with his right arm while wearing the 
orthosis (B) than without it (A). The arrow maiks 
the midline of his body. The right-most row of 
targets was beyond his teach. 




RESNA’94 • Jane 17-22, 1994 



166 



THE ANALYTICAL INADEQUACIES OF TREADMILL- MOUNTED FORCE PLATFORMS 

Ken J. Stewart*, Rory A. Copper^-^, Rick N. Robertson* 

Human Engineering Laboratory Department of Rehabilitation Science and Technology 

California State University School of Health and Rehabilitation Sciences 

Sacramento, CA 95819-6019 University of Pittsburgh, Pittsburgh, PA 15261 



This paper presents a critical review of a 
method for determining the maximal vertical 
force while walking on a force plate mounted 
on a treadmill by R. Kram and AJ. Powell*". 
The purpose of this device is to present the 
clinician with a viable means of measuring 
the maximum ground reaction force of an 
ambulatory subject over a number rf 
consecutive limb contacts. With this design, 
however, the anterio-posterior (A-P) and 
medio-lateral (M-L) forces cannot be 
measured, leaving only the superio-inferior 
(S-I), or vertical, reaction force as the sole 
parameter to be studied. We believe that the 
estimation techniques implemented by the 
authors does not justify the neglect of the 
horizontal forces. We propose an alternate 
estimation technique but suggest that a 
hardware solution can resolve the three 
orthogonal forces. 



Design 

Kram and Powell's design consisted of a motorized 
treadmill fabricated from standard parts. A 
rectangular section was cut out of the middle of the 
treadmUl to accommodate the mounting of the force 
platform. The force platform was positioned within 
the treadmill so that the top of the platform was 
flush with the supporting surface of the treadmill. 
This was done so that the belt would ride smoothly 
across both the treadmill and force platform and not 
exert any significant S-I forces onto the platform. 

The force platform mounted into the treadmill 
structure was a commercially available AMTI 
(Advanced Mechanical Technologies, Inc.), model 
OR6-5-1. This platform is 1.21m long and 0.46m 
wide which is longer than the standard force 
platform. This size of plate was used to 
accommodate the sliding of the stance foot across 
the treadmill as the belt moves. According to 
Rosenrot et al.*^ the contact distance (D^) for a 
walking adult male is calculated from the following 
relationship: 

Dc = 0.665 + 0.25V 

'• ) O 



where V is the velocity of the belt. This allows this 
system to record the ground reaction forces (rf 
w allrin g speeds up to ~2m/s. For a running adult 
male Munro et al.*^ empirically determined D^ to 
observe the following relationship: 

Dc = 0.530 + 0.095V. 

From this relationship the design is useful for 
running speeds up to ~7m/s. 

The type of motor used (electric, gas, pneumatic, 
hydraulic, etc.) was not indicated. To reduce the 
effects of vibration noise, the motor was mounted 
next to the treadmill and aligned to drive the front 
drum through a rubber coupling. 

Discussion 

The deagn of this system does not allow for the 
measurement of the A-P and M-L forces. Assuming 
the belt is relatively stiff and the coefficient of 
friction between the belt and platform is low, any 
true horizontal forces applied to the belt above the 
force platform will not be seen by the platform but 
rather at the drums at either end of the treadmill. It 
is the drums that the belt is attached to, not the force 
platform. Any horizontal forces that are seen at the 
force platform are the result of vertical forces 
generating friction between the belt and the 
platform. The A-P reaction force that the force 
platform measures changes due to the moving belt 
and the variable friction that is being applied to the 
force platform as the S-I force changes. 

The authors understand that horizontal forces 
cannot be measured with their design so they put 
their efforts into the justification that these forces 
are unnecessary. To justify the neglect of the 
horizontal forces, the vector product of the S-I and 
A-P forces was calculated. Rather than using actual 
signals to calculate the angle diffaence between the 
vector product and S-I component, estimations of 
the S-I and A-P forces were used. Kram and Powell 
estimated the S-I signal as 1/2 of a sine wave as 
shown in Frgure 1. Also in Figure 1 is S-I force 
<tata for a walking subject that was measured using a 
standard floor mounted force platform. 



RESNA ’94 • June 17-22, 1994 



167 



ANALYTICAL INADEQUACIES 




Figure 1. Single term sinusoidal estimation vs. 

measured vertical ground reaction force. 



As can be seen by the plot in Figure 1, the sinusoidal 
estimation of the S-I force is not adequate. There is 
a significant amount of deviation from the measured 
signal. 

Kram and Powell’s estimation of the A-P signal is a 
fiill inverted sine wave at 1/10*^ the amplitude of 
the S-I signal estimation. This estimation is shown 
in Figure 2. Also shown in this figine is a typical 
A-P force signal that was measured from a walking 
subject using a standard floor mounted force 
platform. As can be seen from the plot, the 
sinusoidal estimation fw the A-P force signal is also 
unaccqrtable. Although this estimation is better 
than that of the S-I component, neither can be 
construed as viable estimations. 




Figure 2. Single term sinusoidal estimation vs. 
measured A-P ground reaction force. 



Figure 3 shows the vector product for the signal 
estimations that Kram and Powell used fw their 
calculations. Figure 4 shows the vector products (rf 
the data that were measured using a standard floor 
mounted force platform. 




Figure 3. Vector product for the single term 

sinusoidal estimations every 1/10*^ of 
the contact time. 



Figures 3 and 4 are scaled the same, making it fairly 
easy to observe that there are significant differences 
in the vector products from the estimated signals to 
the measured signals. 




Figure 4. Vector product for the measured ground 
reaction forces every 1/10^^ of the contact 
time. 



The magnitude of the vector product is a maximum 
of -1.5% larger than the S-I magnitude. Kram and 
Powell calculate the vector product, using the 
estimations of the signals, to be a maximum of 1 1 .3° 
from vertical. The vector product of the measured 
data was found to be a maximum of 30.6° from 
vertical. This is a profound difference (170.8%) 
and, therefore, invalidates their basic assumption 
that the horizontal forces can be neglected. 

If it is necessary to make estimations of these 
signals, it would be much more accurate to use a two 
term Fourier Series estimation. Figure 5 illustrates 



the mcreased accuracy of a two term sinusoida 
estimation of the vertical reaction force. 



RESNA *94 • June 17-22, 1994 



168 



ANALYTICAL INADEQUACIES 




Figure 5. Two tCTin sinusoidal estimation vs. 

measured v^cal ground reaction force. 



The equation used far this estimation, Ey is as 
follows: 

Ey = I a sin(jit) + a/ 2.75 sin(3jit) I 

where: 

a = Fy.^o .95 and t=o-»i. 

A two term sinusoidal estimation for the anterio- 
posterior forces is shown in Figure 6. The equation 
for this estimation is as follows: 



E, = -a/7 sin(27it) - a /20 sin(47it). 




Figure 6. Two tom sinusoidal estimation vs. 
measured A-P ground reaction force. 



The vector product of the vertical and anterio- 
posterior faces, shown in Figure 7, more highly 
corresponds to that of the measured faces. The 
correlation coefBcients'^ of the measured forces to 
the single and two term estimations fa bofli the 
vertirai and anteria-postaiOT forces are as follows : 



Estimations 


Vertical Force 




Sinele Term 


0.662 


0.908 


Two Term 


0.983 


0.973 




contact time. 

CmiclusioD 

Most biomechanists would agree that the usefulness 
of ground reaction forces are the subsequent 
calculations of net joint forces and muscle moments. 
This can only be done if the horizontal forces are 
accurately measured (at least the A-P forces for two 
dimensional analysis). 

One ^jproach to this problem is the instrumentation 
(rf the legs of a treadmill. If the force transducers 
are located on the legs, any force that is ^lied to 
the treadmiU will be sensed and measured. These 
transducos must be triaxial (biaxial for 2-D 
analysis) to accommodate the measurement of all 
three force and three moment components. The 
treadmill must also be custom designed to maintain 
a high system natural fiequency (>150 Hz). If the 
syston natural frequency is too low, the measured 
force signal will be distorted. The speed of the belt 
must also be measured continuously so that the 
center of pressure can be adjusted appropriately. 

References 

1 R. Kiam and AJ.P«»weU. A Treadmill-Mounted FbiwPlatfom^ 

J. of Applied Physiology, 6m: 1692-1698, 1989. 

2. Roseniot, P., J.C Wall, and J. CJiaiteris. Tta relationship 
between velocity, stride, time, support time and swing time during 
normal walking. J- Hunt, Mov. 6:323-325, 1980. 

3. Monro, CF., D.L Miller, and AJ. Puglevand. Ground reaction 
forees in tunning: a reexamination. J . Biomech. 20:147-155, 1987. 

4. The Student Edition of Matlab, Prentice-Hall, 1992. 

Ken Stewart 

Human Engineering Laboratory 
California State University 
Sacramento, CA 95819-6019 



COPY AVAILABLE 



RESNA’94 • 




7-22, 1994 



169 



ACCURACY OF UPPER-EXTREMITY MOTION 
BY HEALTHY AND DISABLED SUBJECTS 

Cameron N. Riviere 1, Nitish V. Thakor^, and Harry F. Rizer^ 

Departments of Mechanical Engineering! and Biomedical Engineaing^, Johns Hopkins University 

Maryland Rehabilitation Center^ 

Baltimore, MD U.S.A. 



ABSTRACT 

The effects of frequency and type of motion 
on attainable levels of accuracy in computer int^ace 
use are studied. Frequency limitations on accurtK:y in 
one-dimensional motion are presented for three 
subject groups. The maximum limiting frequency for 
accuracy of upper-extremity motion is found to be 2 
Hz. The relative performance of three commercially 
available pointing devices is presented for two subject 
groups. A pen-type interface is found to offer the 
best results for healthy subjects, but the worst results 
for persons with movement disoMers. 



BACKGROUND 

The upper-extremity dexterity of human 
users, i.e., coorcUnation and speed of motion, is an 
important consideration in the design of teleoperative 
systems for rehabilitation Q). Determination of the 
intended command input when noise is present is a 
matter of concern to rehabilitation engineers Q). It is 
clearly desirable for a subject to operate a computer 
under conditions which allow him to use the type of 
bodily motion, and the frequency of motion, which 
offer the greatest accuracy. Previous studies have 
suggested that no single approach is best for all users 
(2), but it may be that certain general principles can 
be found. 

The function of assistive computer interfaces 
is effective transduction of the user’s wishes; that is, 
the person knows what he wants, but needs to 
communicate this to the computer. We have 
developed motion tasks which are simple and easy for 
usere to anticipate, in order to study the effectiveness 
of interfaces to transmit subjects’ commands to the 
computer. We deffne the Accuracy Index, a measure 
of the degree to which the subject stays with the 
target throughout the test. With the Accuracy Index 
we study the limitations which movement frequency 
places on continuous motion by healthy and disabled 
persons. 

The assumption that the type of motion 
required by an interface is one that offers the user a 
good level of control accuracy is implicit in a 
person’s use of an interface. In order to determine 
how attainable accuracy levels vary with different 
types of motion, we study three different pointing 
devices (mouse, pen, and trac-ball) and compare the 
relative error levels generated during the use of each. 



RESEARCH QUESTIONS 

1. What limitation does frequency of motion place 
on accuracy of motion by humans, particularly for 
rehabilitative computer int^aces? 

2. Do Afferent commercially available interfaces 
present significant differences in the levels of accuracy 
attainable by (a) healthy persons, and (b) persons with 
tremcH? 

METHOD 

Two experiments are presented. 

Frequency experiment: A small point-like 
target oscillated sinusoidally along a vertical line 
segment on a computer screen. The subject moved a 
standard Microsoft Mouse, attempting to keep the 
cursor pointing at the moving target For each test, 
the mean-squared error (MSB) was calculated* 




where t and o are the target and human output 
motions, respectively, ^ is a time index, and N is the 
total number of data samples in each test. The root- 
mean-squated oror, 

RMSE = iMEE , ( 2 ) 

was then calculated. The RMS target amplitude was 
also calculated: 

^RMS = tl . (3) 

The level of accuracy with which each subject 
generates the desired motion is represented by a 
measure called the Accuracy Index (AT), which we 
define as 

AI=I.RMSE 

?RMS 

An A/ of zero indicates that the RMS error is the 
same as if the cursor had been left untouched in the 
center of the screen. Perfect motion generation 
results in an AI of unity. 

Three groups of subjects completed the 
experiment: the young group (eight healthy subjects, 
age 23 + 3); the old group (four healthy subjects, age 
72 ± 1); and the tremor group (four tremor patients, 
age 65 ± 16). 

^ Device Comparison: A target moved along 
the vertical line according to a broadband signal. The 



RESNA ’94 




une 17-22, 1994 



170 



Accuracy of UppCT-extremity Motion 



subject again attempted to keep the screen cursor 
pointing at the target. This experiment was done 
with a Microsoft Mouse, a Logitech Trackman trac- 
ball, and an Appoint MousePen (a pen*shaped 
mouse). The MSE (see equation (1)) was calculated 
and us^ to compare the devices. 

This experiment was completed by five 
healthy subjects (age 27 ± 13) and two subjects with 
tremor (age 58 ± 21). 




(a) 




(b) 




(c) 

Figure 1. A/ for frequency experiment, (a) Young 
group (8 subjects). 0>) Old group (4 subjects), (c) 
Tremor group (4 subjects). Plots indicate mean ± 
standard deviation. 



RESULTS 

Frequency experiment: Figure 1 shows the 
AI results for the 1-D experimenL llie results for the 
young group show a marked drop in accuracy 
beginning between 1.7 and 2.0 Hz, crossing zero at 
2.0 Hz. The results for the old group cross zero at 
1.6 Hz, and those for the tremor group cross zero 
much sooner, at 0.5 Hz. Therefore, above 2 Hz, the 
accuracy of all groups is worse than if they had not 
moved the cursor at all. 

Device comparison: Figure 2 shows the 
results of the device comparison task. Each subject’s 
results are normalized by the subject’s mouse MSE, 
so the mouse results always appear as unity. Four of 
the five healthy subjects achieved their best results 
(i.e., lowest MSE) with the pen, and all five did 
better with the pen than With the mouse. The two 
tremor subjects both did considerably worse with the 
pen than with the mouse. The one tremor subject 
who completed the task with the trac-ball did much 
worse with the pen than with the trac-ball. 




Subject 

Figure 2. MSE results of the device comparison tesL 
Results for each subject are normalized to that 
subject’s mouse MSE. N1-N5 are the healthy 
subjects, and T1 and T2 are the tremor subjects. (The 
tremor subject T2 did not complete the trac-ball task.) 

DISCUSSION 

Frequency experiment: The data confirm the 
expectation that the young group would yield the best 
accuracy results. The young group is ^so the most 
consistent in accuracy. For the old group, the AI 
curves are similar in shape to those for the young 
group, but their results are slightly lower. Results 
for the old group are generally less consistent than for 
the young group. As expected, the tremor group 
yielded the worst AI results. 

The AI results for the young group suggest 
that the accuracy limit, or bandwidth, for upper- 
extremity motion generation is 2 Hz, in that AI was 
less than zero above this point. For the old group 
and the tremor group, the AI zero-crossing frequencies 

185 

June 17-22, 1994 



RESNA ’94 



171 



Accuracy of Upper-extremity Modem 

are always less than 2 Hz. It then seems that 2 Hz is 
a maximum value for the accuracy limit, which may 
decrease due to factors such as age and movement 
disorders. We would expect this value to be no less 
than 1 Hz, since Mann et al. (4) reported that the 
average p^ominant frequency for everyday tasks is 1 
Hz. We also would not expect this value to be 
considerably greater than 2 Hz, since Elkind and 
Sprague (^.studying pursuit tracking using 
rectangular taiget spectra, reported that signal-to-noise 
rados decreased sharply when target baitdwidths w^ 
above 1 Hz, and informadon rates at 2 Hz w^ half 
those at 1 Hz. 

Device comparison: Larger-scale study is 
needed to derive conclusive results from the device 
comparison, but it appears that the pen is the best 
interface tested for accurate modon by healthy 
subjects. The results for the tremor subjects suggest 
that movement disorders such as tremor not only 
decrease subjects’ overall accuracy of movement, but 
also alter the accuracy of diff^nt types of movement 
reladve to one another. The difficulty of pen 
manipulation evident here shows the need many 
persons with tremor have for assisdve handwridng 
devices or subsdtutes for handwriting. 

CONCLUSION 

The limidng frequency, or bandwidth, for 
upper-extremity motion by healthy persons is 
approximately 2 Hz. Factors such as age and 
movement disorders may decrease this value. 
Frequency limitadons on accuracy of upper extremity 
modon may limit the p^ormance of man-machine 
interfaces such as assisdve computer int^aces for the 
disabled. 

Movement disorders can be seen not only to 
decrease overall modon accuracy, but also to change 
the reladve levels of accuracy attainable by diffe^nt 
types of modon. While healthy subjects generated 
modon most accurately with a pen, it was the worst 
device for accuracy by tremor subjects. A significant 
need therefore exists for devices to aid or supersede 
handwriting by persons with tremor. 

REFERENCES 

(1) K. Behbehani, G. Kondraske and J. Richmond, 
"Invesdgadon of upper extremity visuomotor control 
performance measures," IEEE Trans. Biomed. Eng., 
vol. 35, pp. 518-525, 1988. 

(2) A. Downing, B. Martin and L. Stem, “Methods 
for measuring the characterisdes of movements of 
motor-impaired children,” Assistive Technology, vol. 
2.4, pp. 131-141, 1990. 



(2) P. Riley and M. Rosen, “Evaluadng manual 
control devices for those with tremor disability,” /. 
Rehab. Res. Dev., vol. 24, no. 2, pp. 99-110, 1987. 

(i) K. Marm, F. Werner and A. Palmer, "Frequency 
spectrum analysis of wrist modon for activides of 
daily living," /. Orthop. Res., vol. 7, pp. 304-306, 
1989. 

(5) J. Elkind and L. Sprague, "Transmission of 
informadon in simple manual control systems," 
IEEE Trans. Human Factors Electron., vol. HFE-2, 
pp. 58-60, 1961. 

ACKNOWLEDGMENTS 

Funding for this project was povided by the 
Nadonal Insdtute on Disability and Rehabilitadon 
Research (grant number HI 33G30064). 

Thanks to the test subjects; to Dr. S.G. 
Reich for providing access to subjects; and to K. Ng 
and M. Fung for preparing data. 

Cam Riviere 

Department of Mechanical Engineering 

Johns Hopkins University 

Baltimore, MD 21218 USA 

(410)516-6570 

FAX: (410)516-7254 

e-mail: cam@jhuvms.hcf.jhu.edu 



RESNA ’94 • June 17-22, 1994 



172 



AN INSTRUMENT FOR TESTING AIMED MOVEMENTS AND APPLIED FORCES 



Steven A. Garand M.S., Albert M. Cook Ph.D., Miriam LeGare Ph.D., and Thomas L. Grey M.S. 
Biomedical Engineering Program, California State University 
Sacramento, CA U.S.A. 



ABSTRACT 

Two disadvantages of systems for studying aimed 
movements are the lack of hardware versatility 
and their nonacconunodation to the physical 
limitations of motorically disabled subjects. A 
general automated system was designed and 
fabricated to overcome these limitations. The 
design goals were to provide a user 
programmable means of generating an audible 
tone and multiple types of target displays (shape, 
position, and timing) on a working surface (15.5 
X 23.5 inches) Much is adjustable from 0° to 
90°. The system's work surfece needed to 
provide a means of recording the force applied, 
the position of touch and a remote optical device 
(head pointer) with a spatial resolution of 0.5 
inches. The results of the tests with a disabled 
subject show that the system meets these goals. 



BACKGROUND 

Specialized equipment has been developed for 
studying the aimed movements of able-bodied 
humans and nonhuman primates. The 
instrumentation is usually designed to answer 
questions derived from specific theoretical 
frameworks which limits the adaptability of the 
instruments. 

For instance, Soechting et al. (1) used 
instrumentation to derive a model of joint angle 
movements in space. The Subject moved the 
hand in geometric shapes in free space. The 
position of the elbow and of the wrist in space 
was obtained by a pair of ultrasound emitters and 
a set of three orthogonal, linear microphones. In 
addition, elbow angle of flexion-extension was 
measured goniometrically. Georgopoulos et al. 
(2) used a system to investigated the factors that 
influence the spatial characteristics of aimed 
movement trajectories of male rhesus monkeys. 
The central 9.8 by 9.8 inches of a plane surface 
served both as a display and as a working surface 
over wiiich the animal moved a manipulandum. 
Nine light-emitting diodes (LED's) were used as 
targets. The manipulandum was a handle that 



the animal grasped and moved over the working 
surface. Shaft angle encoders measured the two 
angles of the manipulandum. 

Problems with using the instrumentation 
previously described in studying subjects who are 
motorically impaired include the following. The 
first system (1) was unable to display targets; it 
used a goniometer and emitters which added load 
to the subject's arm. The second system (2) was 
limited to displaying nine target locations of a 
fixed shape and required a subject to grasp a 
manipulandum. Neither system measured 
applied force, provided for remote target 
manipulation, produced an audible tone, sensed 
the condition of an external switch or provided 
for user programming. 

STATEMENT OF THE PROBLEM 

If aimed movements are to be studied in motor 
disabilities, there are some constraints on the 
type of equipment Miichr can be used. The 
system must not interfere with the subject's 
movements either by adding load or restricting 
arm motion in any plane. It should not require 
consistent grasp capability or prolonged pressing 
of a switch. Further the target array must present 
targets of any sh^, at any position, and at a 
time defined by a user generated program. No 
system was found that could meet all of these 
design requirements. 

RATIONALE 

In order to overcome the problems in using 
existing system designs to study and assess aimed 
movement, a unique system was developed (3). 
The goals for this system are given as follows. 

The system had to display targets. The system 
had to automatically control both the time at 
which a target is presented and the position of 
the target on the working surface. Geometric 
arrays (including alphabetical characters) had to 
be available as targets. 



187 



RESNA’94 • June 17-22, 1994 



173 



Testing Aimed Movements and Forces 



The system had to sense an object (e.g., 
arm/hand) on a two dimensional plane without 
adding load to the arm/hand. The spatial 
resolution of the position sensing system had to 
be at least 0.5 inches with a temporal resolution 
of 20 ms (sampling rate of 50 liz). The work 
area had to be at least 15.5 inches high and 23.5 
inches wide and had to be adjustable from 0° 
(horizontal) to 90® (vertical). The working 
surface had to sense an applied force. The 
targets on the working surface had to be enable 
of activation by a system using a remote optical 
device (head pointer). 

The status of an external switch had to be 
monitored to provide for manual target 
activation. All relevant events, including 
position of contact, condition of switch, applied 
force, and head pointer position had to be 
recorded on disk for future analysis. 

The instrument had to be capeble of generating 
an audible tone as well as sending seven bit 
ASCII information to a computer input 
emulating interface. The system had to be 
programmable by the user and it had to perform 
self diagnostics and report any errors to the user. 

DESIGN 

The system consists of an instrument, a host 
computer, host computer software and an 
optional computer containing a computer input 
emulating interface >^4iich allows the instrument 
to send signals to the computer as if the signals 
were typed from the computer keyboard. 

The instrument consists of a display assembly, 
force measurement system, an instrument 
computer, power supplies, an interface cable, a 
communication (RS232) cable, an external 
switch, an optical head pointer (which allows the 
display assembly to determine the direction in 
which the head pointer is aimed), and system 
software. 

The display assembly consists of a matrix of 
1,536 LED's and the associated electronic 
steering driving circuitiy. The force on the 
working surface is determined by measuring the 
sum of the signals from three bipolar strain 
gauges which support the working surface. The 
instrument computer is based on an Intel 8088 
CPU running at a clock speed of 8 MHz. The 
position of an object on the working surface is 



determined by measuring the state (blocked or 
not-blocked) of infrared emitter detector pairs 
positioned around the perimeter of the work 
surface. The optical head pointer position is 
determined by sequentially illuminating the 
LED's on the work sur^e while measuring the 
response of the photo diode in the head pointer. 

The function of the system is to present a 
stimulus in the form of an illuminated target or 
an audible sound and to record the pertinent data 
of a response in the form of a position profile on 
the work surface, a force applied to the work 
surface, and/or a logged event. A text file located 
on a host computer controls the way in which the 
system operates. 




SYSTEM LAYOUT 



Two sets of software were written for the system. 
The first set operates the host computer and the 
second set operates the instrument controller. 
The host computer software provides the 
interface for the user of the system. It translates 
a program file into a set of binary arrays that can 
be interpreted and executed by the instrument 
software. It also translates the data the 

instrument has recorded into an ASCII text file 
which is sent to the screen (in diagnostic mode) 
or to a disk drive following a program execution. 
The host software is written in C. 

The instrument software operates the instrument 
as defined in the user generated program file. 
This software records the data and transfers the 
data to the host software. The instrument 
software also has a set of routines to perform a 
self diagnosis of the instrument. The instrument 



Er|c 174 




RESNA’94 • Ju^l7-22, 1994 



Testing Aimed Movements and Forces 

software resides in the instrument ROM and is 
written in ANSI C and assembly language. 

A user generated program file defines how the 
instrument is to operate. The program file is an 
ASCII text file that is generated from a text 
editor such as WordPerfect. Data manipulation 
programs, written in C. were also developed to 
extract specific information, e.g., times between 
events, fi"om a data file stored on a disk drive. 

DEVELOPMENT 

The original system design was laid out in a 
schematic capture program. System modules 
were assembled on wire wrapped boards and 
tested individually. Problems with the design 
such as an insufficient range of the position 
sensing system were corrected and the schematics 
were revised. Artwork was generated fl'om the 
schematic files and printed circuit boards (PCB's) 
were fabricated fi"om the artwork. The system 
hardware was assembled and tested. Problems 
with the hardware such as faulty wire 
connections were corrected. The System 
Software was then developed on an IBM PC. 
Final testing and modification to the integrated 
software and hardware system were completed 
with the use of an 8088 emulator. 

EVALUATION 

The system was tested to determine if the system 
gocds were met. The system was tested with one 
subject, an adult female with a motor disability. 
The three types of tests performed were target 
display and position sense tests, a head pointer 
test, and a force sense test. The data fl’om the 
pilot study tests were stored on a disk at the end 
of each experiment by the system host computer. 
The data on this disk were analyzed by an IBM 
PC using three data manipulation programs. 

The results of these tests showed that a 1 inch 
square target was displayed following a switch 
closure. The system automatically controlled the 
time the target was presented and the position of 
the target. The first three letters of the alphabet 
were also displayed on the working surface. 
Making contact inside the target area caused the 
target to turn off. a marker number was recorded, 
and a momentary audible tone was turned on. 
The system sensed the position of a foam pad on 
the working surface in x and y coordinates with a 
spatial resolution of 0.5 inches and a temporal 



resolution of 30 Hz. The proper signals on the 
computer interface cable were verified on an 
oscilloscope. Position contact data, the status of 
a closed switch, the force applied to the working 
surface, and head pointer position were recorded 
on disk. A system self test was also successfully 
performed. 

DISCUSSION 

The results of the three types of tests performed 
in the pilot study show that the system meets the 
required design goals. The system has recently 
been used in a study of adaptation to the doll 
reflex in wheel chair athletes. The system should 
be usefiil in both basic and clinical research and 
in assessing the physical skills necessary for 
augmentative communication. 

REFERENCES 

(1) Georgopoulos, Apostolos P.. Kalaska, John F. 
and Massey, Joe T. "Spatial Trajectories and 
Reaction times of Aimed Movements; Effects 
of Practice. Uncertainty, and Change in 
Target Location." Journal of Neurophysiology 
46(1981): 725-743 

(2) Soechting, J. F., Lacquaniti, F. and Terzuolo, 
C. A. "Coordination of Arm Movements In 
Three-Dimensional Space. Sensorimotor 
Mapping During Drawing Movement." 
Neuroscience 17 (1986): 295-311 

(3) Garand, S.A. "An Instrument for Testing 
Aimed Movements and Applied Forces." 
Masters Thesis, Biomedical Engineering 
Program. California State University, 
Sacramento (1993) 

ACKNOWLEDGMENT 

This project was developed in association with 
the School of Engineering and Computer 
Science, California State University, Sacramento. 



Steven Garand 

Design Engineer 

Techmed Consulting 

2419 Stokewood Way 

Rancho Cordova, CA 95670 U.S.A. 

(916)631-4145 

FAX (916) 631-4145 



189 

RESNA ’94 • June 17-22, 1994 



175 



THE COMPARATIVE EVALUATION OF ISOTONIC AND ISOKINETIC MODES OF TESTING WITH 
ERGONOMIC AND REHABILITATION PERSPECTIVE 

Kinda A. Khalaf, M.S., Mohamad Pamianpour, Ph. D., and Patrick Spaito, B.S. 

Biodynamics Laboratoiy, The Ohio State University, Columbus, Ohio 43210 



Abstract 

The climbing costs of musculoskeletal 
injuries, and the implementation of the Americans 
with Disabilities Act create an increasing need for 
ergonomists and clinicians to develop effective 
means to assess the feasibility of performing a task. 
It is important to develop a unified approach to 
quantification of job demand and fimctional 
capability profiles. The objective of this work was to 
develop a data base of fimctional capability in terms 
of isokinetic and isotonic strengths for multiple 
joints. Ten subjects performed a variety of multi- 
joint strength and speed tests in the isometric, 
isokinetic and isotonic modes. The results show that 
the isokinetic mode is a more systematic and 
efficient method of characterizing the strength 
c^)ability. The tradeoff between efficiency of tests 
and simulation of fimctional attributes of physical 
activity requires fiuther exploration. 



Introduction 

Musculoskeletal injuries, of which low back 
pain constitutes the largest portion, are the second 
most firequent cause of worker disability, following 
cardiovascular diseases. According to the bureau of 
labor statistics, back injuries account for 
approximately 32% of all occupational disability 
injuries and consume up to 42% of all compensation 
payments (1). Ergonomists and researchers look for 
ways to quantify the feasibility of performing certain 
job tasks as related to individual fimctional 
capability. The Americans with Disabilities Act 
(ADA) Of 1990 dictates that employers must provide 
reasonable accommodations, which are not an xmdue 
hardship, so that any individual with disability can 
perform the essential function of a job. Potentially 
qualified disabled individuals must undergo a series 
of medical examinations to demonstrate the ability to 
perform the required job fimction. This creates a 
need to develop effective means to assess feasibility 
of performing a task by a unified approach to 
quantification of job demand and fimctional 
capability profiles. 

Ergonomic evaluation of task demands 
could be used to provide the required torque and 
velocity at specific joint positions. This may be 



obtained from biomechanical multi-link models of 
lifting tasks. (2), (3), (4). The question whether a 
subject/patient (an applicant for a given job or an 
injured worker) can perform the analyzed task in 
terms of its basic elements of performance (i.e. 
strength, range of motion, endurance, speed, etc.), is 
the fundamental challenge in rehabilitation 
engineering. 

In his paper about muscular performance 
evaluation, Sapega (S) reviews the evaluation of 
fimctional performance based on isometric or 
isokinetic modes of testing. Despite the fact that 
isometric and isokinetic test modes are known to be 
less functional as compared to dynamic daily 
activities, th^ have been extensively used in clinical 
evaluation (6). Up to date, there has been no stu^ 
that compares the feasibility of performing a specific 
task in isokinetic and isotonic modes. Therefore, the 
objective of this stu^ is to develop a data base of 
fimctional cjqrability in terms of isokinetic and 
isotonic strengths for multiple joints. Additionally, 
the psychometric attributes of these two dynamic 
modes of testing will be compared. 



Methods 

Ten healthy male normal subjects 
participated in this study. The mean (s.d.) age, mass, 
and stature of the subjects were 26.2 (3.8) years, 83. 1 
(14.0) kg, and 178.6 (10.7) cm, respectively. The 
strength and speed of the subjects were studied for 
each of the isolated joints: elbow, shoulder, back, 
hip, knee, and ankle using the KIN_COM 125E Plus 
muscle testing and training system from Chattecx. 
corp. Each joint was tested using the isometric, 
isokinetic, and isotonic modes of testing. In the 
isometric testing mode, the subjects were asked to 
use maximum exertions at hold angles covering the 
range of motion for both extension and flexion. 
During the isokinetic mode, the subjects were 
instructed to perform maximum exertion at each 
specific set velocity. These speeds were selected 
based on the values from literature as well as the 
results of the coordinated lifts performed in a 
previous stu^ (3). During isotonic tests, the joint 
velocity and torque were measured while the 
resistances were fixed. Both the velocities and the 



176 



RESNA ’94 • June 17-22, 1994 



THE COMPARATIVE EVALUATION... 



Table 1. The mean, maximum, and minimum measured torque during isotonic knee extension at specific 
joint positions and velocity range. 



Velocity (deg/s) 


Positions (deg) 


110-119 


120-129 


130-139 


140-149 


150-159 


160-169 


170-180 


0-25 


14 • 












11 




171.1 * 


N/A 


N/A 


N/A 


N/A 


N/A 


20.8 




80.0 t 












86.0 




267.0 0 












71.0 


25-75 


53 












41 




195.2 


N/A 


N/A 


N/A 


N/A 


N/A 


75.8 




98.0 












20.0 




296.0 












181.0 


75-125 


56 


2 










30 




203.4 


241.0 


N/A 


N/A 


N/A 


N/A 


142.6 




102.0 


235.0 










70.0 




294.0 


247.0 










217.0 


125-175 


14 


44 


3 








18 




188.7 


222.8 


247.7 


N/A 


N/A 


N/A 


177.6 




117.0 


136.0 


245.0 








13410 




294.0 


291.0 


252.0 








228.0 


175-225 




24 


3 








21 




N/A 


223.0 


247.7 


N/A 


N/A 


N/A 


158.1 






136.0 


245.0 








92.0 






291.0 


252.0 








219.0 


225-275 


1 


27 


6 


2 






12 




168.0 


214.4 


259.7 


261.0 


N/A 


N/A 


175.0 




168.0 


145.0 


226.0 


261.0 






152.0 




168.0 


291.0 


292.0 


261.0 






219.0 



Table 2. The mean, maximum, and minimu m measured torque during isokinetic knee extension at 
specific velocity and range of joint position. 



Velocity (deg/s) 


Positions (deg) 




110-119 


120-129 


130-139 


140-149 


150-159 


160-169 


170-180 


10 


179.6 * 


203.7 


204.5 


204.0 


208.2 


191.6 


161.6 




76.0 T 


159.0 


171.0 


177.0 


157.0 


158.0 


109.0 




251.0 0 


250.0 


243.0 


254.0 


253.0 


235.0 


164.0 


50 


102.2 


151.5 


174.4 


195.4 


220.0 


227.8 


170.7 




46.0 


93.0 


143.0 


166.0 


188.0 


210.0 


105.0 




160.0 


187.0 


205.0 


219.0 


251.0 


258.0 


211.0 


100 


81.9 


110.2 


133.1 


150.5 


168.9 


193.1 


168.8 




43.0 


80.0 


110.0 


129.0 


144.0 


164.0 


105.0 




108.0 


140.0 


158.0 


182.0 


200.0 


230.0 


198.0 


150 


60.2 


94.9 


127.8 


146.9 


168.1 


177.6 


152.2 




27.0 


54.0 


116.0 


113.0 


114.0 


137.0 


96.0 




97.0 


116.0 


148.0 


178.0 


228.0 


224.0 


194.0 


200 


71.8 


125.4 


172.1 


199.1 


218.5 


215.9 


146.0 




43.0 


48.0 


110.0 


148.0 


159.0 


183.0 


121.0 




91.0 


177.0 


216.0 


251.0 


269.0 


247.0 


159.0 


250 


132.3 


142.0 


114.4 


236.0 


120.0 


112.2 






83.0 


96.0 


93.0 


236.0 


120.0 


98.0 


N/A 




190.0 


200.0 


156.0 


236.0 


120.0 


161.0 





® : Number of observations, * : Mean torque, Miniroiun torque, 0 : Maximum torque, N/A : No 
data were present in these specific cells 



RESNA *94 



191 

• June 17-22, 1994 



OPY AVAILABLE 



177 



resistances used were joint-specific. The knee 
velocities, for example, were 10, 50, 100, 150, 200, 
250 Degrees/s; in the isotoiuc mode, the resistances 
of the knee joint were 10, 50, 100, 150, 200, 200 
Newtons. 

Results 

The descriptive and distribution profiles 
were obtained for each joint. For the purpose of 
illustration, the results are shown based on the 
isotonic and isokinetic knee extension for one 
subject. Table 1 shows the mean, minimum, and 
maximum measured torque (Newton-meter) at 
specific joint position and velocity ranges during 
isotonic mode of testing. Table 2. shows the mean, 
minimum, and maximum measured torque at 
specific isokinetic velocities and position ranges 
during isokinetic testing. 

Figure 1. represents an example of a 
bivariate distribution histogram for e ach 
combination of ranges of joint velocities and 
positions. 



Discussion 

It can be seen finm the tabular and 
graphical results that although a wide range of 
resistances were used during the isotonic exertions, 
the final matrix is rather square as compared to the 
isokinetic data. Hence the proposed question cannot 
be evaluated for those combinations of joint 
velocities and positions that were not represented in 
the data base. This indicates that the isokinetic mode 
of testing in the evaluation of muscular performance 
is a more systematic and efficient method of 
characterizing the strength capability. This study 
does not answer how the isokinetic mode compares 
with unconstrained dynamic activities in terms of 
functionality. 

In conclusion, despite acknowledging that 
the isokinetic mode may not be functional, tradeoff 
between efficiency of tests and simulation of 
fimctional attributes of physical activity requires 
further exploration. Additionally, future studies 
should consider the recruitment patterns dunng 
these modes of testing to farther delineate task 
appropriateness of these tests with respect to 
imconstrained (tynamic tasks (i.e. lifting). 

References 

1. K. L. Grazier, T.L. Holbrook, J.L. Kelsey, and 
R.N. Strauffer, "The fi'equency of Occurrence, 
Impact, and Cost of Musculoskeletal Conditions in 
the United States," The American Academy of 
Orthopedic Surgeons, Chicago, 1988. 




Figure 1. Bivariate Distribution Histogr am of 
isotonic knee extension for one subject 

2. M. Pamianpour, L. Hasselquist, L. Fagan, and A. 
Aaron, "Correlation among isometric, isokinetic and 
isoinertial muscle performance during multi-joint 
coordinated exertions..." European J. ofPhys. Med. 
md Rehab., vol. 3, no. 3, pp. 114-122, 1993. 

3. P.J. Sparto, K.A. Khalaf, and M. Pamianpour, 
"The effect of load, speed and mode of lift on the 
joint energetics during unconstrained lifting and 
lowering activities," Presented at the 1993 ASME 
Winter Annual Meeting, New Orleans, LA, 1993. 

4. P.J. Sparto, M. Pamianpour, and Khalaf, K.A., 
"The reliability and validity of a lift simulator and its 
functional equivalence with fi’ee weight li fting 
tasks," Submitted to IEEE Transaction on 
Rehabilitation Engineering, 1993 

5. J.J. Perrin, V.R. Edgerton, "Muscle Force-velocity 
and power-velocity relationships under isokinetic 
loading", Med. Sci. Sports, vol. 10, pp. 159-166, 
1978 

6. A, A. Sapega, "Muscle performance evaluation in 
orthopaedic practice," The Journal of Bone and Joint 
Surgery, vol. 72a, no. 10, pp. 1562-1574, 1990. 

Acknowledgments 

The authors acknowledge partial support 
from NIDRR, REC grant #H133E30009 and the 
Ohio Bureau of Workers' Compensation, Division of 
Safety and Hygiene. In addition, William Marras, 
Ph.D., and Sheldon Simon, M.D. are gratefully 
acknowledged for their invaluable assistance. 

Kinda Khalaf 
The Ohio State University 
Biodynaraics Lab - Room 140 
Baker Systems and Engineering 
Columbus, OH 43210 



O 



ERJC 178 



192 ? 



iS i' 



r 

k 



RESNA’94 • June 17-22, 1994 



A VALIDATION STUDY OF A LIFT SIMULATOR DURING ISOINERTIAL LIFTS 



Patrick J. Sparto, B.S., Mohamad Pamianpour, Ph.D., and Kinda A. Khalaf, M.S. 
Biodynamics Laboratory, The Ohio State University, Columbus, Ohio 43210 



Abstract 

Dynamometers are being used increasingly 
in the rehabilitation field to quantify the disability of 
injured workers and to train them for their return to 
work. However, the exercises performed with these 
dynamometers have not been shown to be 
functionally equivalent to their real world 
counterparts. Hence no information can be inferred 
from such exercises. A comprehensive set of lifting 
exercises was developed to validate the isoinertial 
mode of the LEDOLift lifting simulator (Loredan 
Biomedical, Inc., Davis, CA). Ten subjects 
performed a variety of submaximal and maximal 
lifts in the isoinertial mode. The data from the 
submaximal lifts were used in a linear regression to 
quantify the behavior of the isoinertial mode. The 
submaximal lifts at higher loads were foimd to more 
accurately represent the theoretical behavior than the 
lifts at lower loads. Overall, the R^ value for the 
measured and predicted loads was .96. When the 
coefficients from the regression of the submaximal 
lifts were used for the regression of the data from the 
maximal lifts, the predicted forces were within 5% of 
the measured forces, R^ = .915. 

Introduction 

The measurement of muscular performance 
has become a necessity in the physical medicine and 
rehabilitation community. Clinicians have used 
dynamometer-based systems for the purpose of 
rehabilitating the injured workers and athletes, and 
for objectively documenting the extent of disability 
for the resolution of legal disputes, the brunt of 
which are Workers' Compensation claims (1). 

Objective quantification of trunk muscular 
performance is most significant since a majority of 
low back pain patients do not present any anatomical 
finding (^. A small portion of patients who become 
chronic are responsible for 80% of the total cost (^. 
Hence, quantitative assessments allow more precise 
evaluation of functional disability and provide 
processes for early return to work. A major cost 
saving effort is directed towards management of 
acute low back pain to avoid the prospect of chronic 
illness (2),(3). 

The importance of an objective means of 
quantifying muscular performance is further 



magnified by the enaction of the Americans with 
Disabilities Act (1991), which requires that muscle 
testing reflect the 'essential function of the task'. In 
addition, companies are obligated to provide 
'reasonable accommodation' for the worker. The 
testing of the functional capabilities of the worker 
should play an important role in the interpretation of 
'reasonable accommodation'. 

Dynamometers are a preferred tool for 
muscle testing and rehabilitation because of their 
safety and objectivity compared to manual testing. 
However, the inference of functional capacity by the 
use of any of these dynamometers has remained 
controversial (i),(4). Consequently, it is essential 
that these machines are proven to be valid in their 
simulation of real world, or free dynamic, activities. 

The LEDOLift (Loredan Biomedical, Inc., 
Davis, CA) is a computer-controlled lift device that 
is supposed to simulate isometric, isokinetic and 
isoinertial lifting conditions. The calibration of the 
electromechanical sensors and the validation of the 
LEDOLift in the isometric and isokinetic modes has 
been shown by Wright et al. (5). However, the 
validation in the isoinertial mode has not been 
reported in literature. The purpose of this paper is 
twofold. First, a model will be developed that 
describes the behavior of the LEDOLift in the 
isoinertial mode. Secondly, the model will be used to 
predict the behavior of the LEDOLift in subsequent 
testing procedures. 

Methods 

To determine the validity of the isoinertial 
mode, ten healthy male normal subjects participated 
in a study after signing a consent form approved by 
the human subjects committee. The mean (s.d) age, 
mass, and stature of the subjects were 26.2 (3.8) 
years, 85.1 (14.0) kg, and 178.6 (10.7) cm, 
respectively. Briefly, each subject lifted and lowered 
a two-handled box attached to the arm of the 
LEDOLift. The simulated masses for the study were 
6.8, 13.6 and 20.5 kilograms (15, 30, and 45 pounds, 
respectively). A linear regression analysis was 
performed to find the best fit model for the measured 
force, based on: 

Fjjj = Cj • m + C 2 • m • a (1) 







u m3* 




i 



RESNA ’94 • June 17-22, 1994 



179 



A VALIDATION STUDY OT A LIFT... 



Table I. Values of CoefBcients and based on linear regression of measured 

force with the sinaulated mass, and mass times the acceleration (Equation I). 



Simulated 

Mass 


COEFFICIENTS 


R2 


Cl 


C2 


6.8 kg 


6.23 


1.18 


0.89 


13.6 kg 


7.92 


1.75 


0.96 


20.4 kg 


8.45 


1.08 


0.98 


All 


8.13 


1.26 


0.96 






Acceleration (tn/s^2) 

■X Regression —4— Theoretical 



Figure 1. The effect of simulated load on the difference between the theoretical force and the measured force 
predicted by linear regression. Lines with (X) are from the regression. Lines with (+) are the theoretical 
forces. Dashed lines represent the 95% confidence interval for the regression line 

. 194 



180 



RESNA’94 • June 17-22, 1994 



BEST COPY AVAIUBLE 



A VALIDATION STUDY OF A LIFT... 



where Fm is the measured force, m the simulated 
mass, a the acceleration of the device arm, and the 
coefficients C| and C 2 to be determined by the 
regression. No intercepts were included in the 
model, hence facilitating a physical interpretation of 
the model. The theoretical force prediction based on 
an isoinertial mode of lift would require c^ and C 2 to 
be equal to 9.81 and 1.0 respectively. Then Cj would 
physically represent the gravitational constant for 
acceleration. The aforementioned regression was 
performed for each simulated mass and for the entire 
data set. A number of other regressions were 
performed that are not presented here. 

Next, maximum isoinertial lifts were 
performed by each subject. The data from these lifts 
were input into the regression model (the right hand 
side of Equation (1)) to determine the predicted 
force. The simulated masses for the maximum 
isoinertial lifts ranged from 33.2 to 95.5 kg. An 
additional regression between the predicted force 
and the measured force from these independent lifts 
was performed. 

Results 

In the regression of the three simulated 
masses, the coefficients Cj and C 2 and of the 
model, fit to each mass separately as well as fit to the 
whole data set, are shown in Table 1. As the mass 
increases, increases. In addition, coefficient Cj 
approaches the value of 9.81 as the simulated mass 
increases. Coefficient C 2 , approaches the value of 
1.0 at the highest simulated mass. 

Figure 1 shows how the regression 
compares to the predicted at each of the loads. The 
marked lines represent the theoretical force (+) and 
the predicted force (x) based on the regression. The 
dashed lines are the 95% confidence intervals for the 
predicted force. As the load increases from 6.8 to 
20.5 kg, the simulated force line approaches the 
theoretical force line. However, the deviation of the 
slopes at 13.6 kg calls for caution against 
extrapolating any information beyond the values 
presented here. The theoretical force, based on c^ 
equal to 9.81 and C 2 equal to 1.00, is within the 95% 
confidence interval for simulated loads of 6.8 and 
20.5 kg. 

For the maximum isoinertial lifts, the 
regression was performed with the coefficients from 
the whole data set (Cj = 8. 13, C 2 = 1.26). When these 
coefficients are used to predict the behavior of the 
LIDOLift during the maximum isoinertial lifts, the 
difference between the measured force and the 
predicted force is approximately 5% ( R^ = .915). 



Discussion 

The results indicate that the simulated mode 
is not exactly isoinertial at lower loads. However, 
the added safety provided by the simulator may 
outweigh this limitation. Clinicians and users of the 
product must account for this when planning a 
testing or training protocol. At higher loads, the 
simulated isoinertial conditions are much closer to 
the theoretical predictions. 

It is intended that this research will lead to 
a cooperative effort between the developers and users 
of similar products, so that the training and testing 
protocols used in the rehabilitation field may 
undoubtedly reflect the activities in real world 
environments. Based on these results, we are 
initiating another study to help setting the simulated 
mass so that the current limitations can be 
compensated. 

References 

1. A. A. Sapega, "Muscle performance evaluation in 
orthopaedic practice," The Journal of Bone and Joint 
Surgery, vol. 72a, no. 10, pp. 1562-1574, 1990. 

2. M. Pamianpour, and J.C. Tan, "Objective quantification 
of trunk performance," Chapter in Back Pain 
Rehabilitation, D'Orazio (editor), Andover Medical 
Publishers, Boston, pp. 205-237, 1993. 

3. M. Pamianpour, L. Hasselquist, L. Fagan, and A. 
Aaron, "Correlation among isometric, isokinetic and 
isoinertial muscle performance during multi-joint 
coordinated exertions..." European J. of Phys. Med. and 
Rehab., vol. 3, no. 3, pp. 1 14-122, 1993. 

4. J.M. Rothstein, R.L. Lamb, and T.P. Mayhew, 
"Clinical uses of isokinetic measurements - critical 
issues," P/iysica/ Therapy, vol. 67, no. 12, pp. 1840-1844, 
1987. 

5. P. Wright, M. Pamianpour, and W. Manas, "Reliability 
and validity of the kinematic and kinetic measures of a lift 
simulator," Proceedings of the Second North American 
Congress on Biomechanics, pp. 529-530, 1992. 

Acknowledgments 

The authors acknowledge partial support 
from NIDRR, REC grant #H133E30009 and the 
Ohio Bureau of Workers* Compensation, Division of 
Safety and Hygiene. In addition, P. Lawrence 
Wright, M.S., William Marras, Ph.D., and Sheldon 
Simon, M.D. are gratefiilly acknowledged for their 
invaluable assistance. 

Patrick Sparto 
The Ohio State University 
Biodynamics Lab - Room 140 
Baker Systems and Engineering 
Columbus, OH 43210 

195 



RESNA ’94 • June 17-22, 1994 



181 



THE EFFECTS OF LOAD, MODE, SPEED OF LIFT ON THE POWER GENERATION, ABSORPTION 
AND TRANSFER DURING A MULTI-LINK COORDINATED LIFTING TASK 

Patrick J. Sparto, B.S., Mohamad Pamianpour, Ph.D., and Kinda A. Khalaf, M.S. 
Biodynamics Laboratory, The Ohio State University, Columbus, Ohio 43210 



I 



I 



Abstract 

The reduction of the large costs associated 
with the chronicity of low back pain is a driving 
force in the development of strategies to rehabilitate 
the injured worker. Quantitative assessment of the 
demands of the job along with the functional 
evaluation of the patient provides a rational basis for 
the process of the return to work. Ten subjects 
participated in a study to determine the power 
requirements of several lifting tasks and to analyze 
the effect of lift parameters on these power demands. 
The hip was found to provide most of the power 
generation in the lifts, while the most power transfer 
occurred in the back. The economy of movement 
was approximately 0.26 throughout the time of lift. 
An ANOVA revealed that a stoop lift is the least 
demanding in terms of econmy of movement. It is 
concluded that the power analysis is a necessary tool 
to complement the information given by an 
examination of external moments, since the former 
naturally reflect the (fynamic parameters of 
performance (i.e. velocity and acceleration) which 
have been implicated as potential risk factors. 

Introduction 

It is recognized that the timely 
rehabilitation of the worker with a lower back injury 
is paramount in the prevention of the chronicity of 
low back pain. Because disability associated with 
chronic low back injury is implicated with a large 
amount of the health care costs (1), it is important to 
develop strategies for rehabilitating the injured 
worker quickly. 

The quantitative assessment of job demands 
and functional testing of disability are needed to 
determine a training protocol that will return the 
injured to work as soon as possible. For example, if 
the biomechanical analysis indicates that a hip 
power generation of 425 Watts is required during a 
lift of 20 kg., and the injured worker is unable to 
meet this task demand, the goal of the rehabilitation 
program is directed to enable the worker by 
strengthening exercises and/or use of optimal lifting 
techniques. 

The analysis of power requirements of the 
joints and the relative shming of power amongst the 



joints in performing a task has been proposed as a 
method for job demand assessment (2). Moreover, 
the patterns of joint generation, absorption and 
transfer produced in a task may lead to a better 
imderstanding of the load sharing during multi-link 
coordinated lift, which may help to distinguish 
between the patient and normal population. 

The purpose of this paper is to present the 
power analysis of a lift and then to detail the effects 
of different job demands on these power 
computations. 

Methods 

Ten healthy male normal subjects 
participated in a study after signing a consent form 
approved by the human subjects committee. The 
mean (s.d) age, mass, and stature of the subjects 
were 26.2 (3.8) years, 85.1 (14.0) kg, and 178.6 
(10.7) cm, respectively. Briefly, each subject 
performed lifts involving three different loads (6.8, 
13.6, and 20.4 kg), three different lifting modes 
(preferred, stoop, and squat), and three different 
lifting speeds (slow, medium, and fast). A more 
thorough description of the lifting protocol and 
analysis of data appears in Sparto et al, (3). The 
computation of power generation, absorption, and 
transfer is given hy Robertson and Winter (4), while 
the economy of movement is described as the ratio of 
power transferred relative to the total power (i.e. 
generated + absorbed + transferred) (2). 

Results 

The results of a lift at 20.4 kg, preferred 
mode of lift, and fast speed for one subject is shown 
in Figure 1 . The pattern of these curves is consistent 
for all subjects. 

The joint positions are shown in Figure 1- 
A, with 0 degrees representing full flexion and 180 
degrees full extension of each joint. The 
hyperextension of the hip and the motion of the 
L5/S1 joint are calculated based on the measured 
value of the knee and trunk angles, using the 
regression equation reported by Chaffm and 
Andersson (5). The tru^ angle is also shown so 
that the motion of the whole body can be better 
described. 



ERIC 






182 



RESNA’94 • June 17-22, 1994 



THE EFFECTS OF LOAD, MODE .. 





knee 

hip 

L5/S1 

trunk 



knee 

hip 

L5/S 1 

box 



knee 

hip 

L5/S1 



Figure 1. The joint position (A), power generation and absorption (B), power transfer (C), and economy of 
movement (D) of a lift of a 20.4 kg. load at a fast speed in a preferred lifting mode by one subject. For the 
joint position, 0 degrees is fully flexed and 180 degrees is fully extended. 



RESNA ’94 



June 17-22, 1994 



COPY AVAILABLE 



183 



THE EFFECTS OF LOAD, MODE .. 



Table 1. The summary results (p-values) of ANOVA, testing the effect of load (L), mode (M), speed (S), and 
their interactions on the energedc parameters during a lifting task. An asterisk indicates significance, a=O.05. 





P-values of the effect of; 


Dependent variables 


L 


M 


s 


LXM 


LXS 


MXS 


Knee power gen. / absorption 


0.040* 


0.0003* 


0.0001* 






0.0001* 


Hip power gen. / absorption 


0.0001* 


0.0001* 


0.0001* 


0.0001* 


0.0001* 


0.0001* 


L5/S1 power gen. / absorption 




0.0001* 




0.0001* 




0.0001* 


L5/SI power transfer 


0.0001* 


0.0001* 


0.0001* 




0.0001* 


0.0005* 


Economy of movement 


0.0001* 


0.0001* 


0.010* 









The patterns of power generation, 
absorption, and transfer, shown in Figure 1-B, and 
1-C should be considered together when analyzing 
the data, since the sum of power at each joint is 
divided into the generation or absorption by muscles 
spaiming the joint, and transfer from the one of the 
segments of the joint to the other. The muscles of 
the knee and hip are generating power throughout 
the whole lilting phase, while generally there is little 
or no transfer. The trunk muscles are initially 
absorbing power and then generating power, while a 
transfer of power from the muscles of the pelvis to 
the back is occurring from 25% of the movement 
time to the end. Additionally, the power transferred 
to the box is shown in Figure 1-B. 

Clinically, a more useful description of the 
data might lie in the economy of movement, an 
example of which is shown in Figure 1-D. The 
value of the economy during most of the cycle is 
approximately 0.26. Meanwhile, the value of the 
economy in the initial portion of the lift should be 
considered with caution due to the small magnitude 
of the power terms. 

The results of the ANOVA of the effects of 
load, mode, speed, and interactions is detailed in 
Table 1. There is a highly significant difference 
between modes of lift for all five variables presented 
here. The results from a Tukey post-hoc analysis 
show the power generation of the knee and hip 
muscles to be less for a stoop lift than the other 
modes of lifts. Additionally, the economy of 
movement is greatest for the stoop lift, while the 
economy of the preferred lifting method is greater 
t han the squat lift. This result is consistent with 
physiological findings (6). 

Discussion 

The complete power analysis described here 
is currently being used as a tool for the assessment of 
the demands of the job. When the power demands 
are compared with a clinically-based strength and 
power evaluation, a viable rehabilitation program 



can be developed. In addition, task analysis and 
functional evaluation protocols based on the power 
requirements reflect the ^mamic components of 
physical activities much better than the 
quantification of external moments alone. 

References 

1. M. Pamianpour, and J.C. Tan, "Objective quantification 
of trunk performance," Chapter in Back Pain 
Rehabilitation, D'Orazio (editor), Andover Medical 
Publishers, Boston, pp. 205-237, 1993. 

2. M. Gagnon, and G. Smyth, "Muscular mechanical 
energy expenditure as a process for detecting potential 
risks in manual materials handling," Journal of 
Biomechanics, vol. 24, no. 3/4, pp. 191-203, 1991. 

3. P. Sparto, M Pamianpour, and K. Khalaf, "The effect 
of load, speed, and mode of lift on the joint energetics 
during unconstrained lifting and lowering activities," 1993 
Advances in Bioengineering, Tarbell (editor), ASME - 
BED, vol. 26, pp. 467-470, 1993. 

4. D. Robertson, and D. Winter, "Mechanical energy 
generation, absorption, and transfer amongst segments 
during walking," Journal of Biomechanics, vol. 13, pp. 
845-854, 1980. 

5. D. Chaffin, and G. Andersson, Occupational 

Biomechanics, Wiley Interscience, 1991. 

6. E. Welbergen, H. Kemper, et al, "Efficiency and 

effectiveness of stoop and squat lifting at different 
frequencies," vol. 34, pp. 613-624, 1991. 

Acknowledgments 

The authors acknowledge partial support 
from NIDRR, REC grant #H133E30009 and the 
Ohio Bureau of Workers* Compensation, Division of 
Safety and Hygiene. In addition, William Manas, 
Ph.D., and Sheldon Simon, M.D. are gratefully 
acknowledged for their invaluable assistance. 

Patrick Sparto 
The Ohio State University 
Biodynamics Lab - Room 140 
Baker Systems and Engineering 
X^lumbus, OH 43210 



ERIC 



184 



RESNA’94 • June 17-22, 1994 



SIG-06 

Special Education 



EDUCATION AND ASSISTIVE TECHNOLOGY: FLORIDA’S MODEL 



Catherine H. George, Sandra R. Osborn 
Assistive Technology Educational Network of Florida 
Orlando, FL 



In response to federal requirements and ongoing 
demands for assistive technology services, the 
Florida Department of Education, Bureau of 
Exceptional Student Education, has implemented 
a statewide plan for providing service, 
technology and training to its entire exceptional 
student school aged population. Built upon a 
foundation of existing specialized support 
services, the state initiative involves four main 
components: regional resource centers, a network 
of local assistive technology specialists, 
a device loan library, and a personal device loan 
system. 

BACKGROUND 

In response to federal requirements and ongoing 
demands for assistive technology services, the 
Florida Department of Education has 
implemented a statewide plan for providing 
technology and training to its entire exceptional 
student school aged population. The Department 
of Education stands as an integral member of a 
cooperative interdepartmental state venture to 
provide a seamless system of service, technology 
delivery and training to all its citizens with 
assistive technology needs. 

OBJECTIVE 



with disabilities. The center, based in 
Orlando, serves as a model for four additional 
resource sites to be established over the next 
five years. ATEN provides assessment, 
technical assistance, training, reference 
library, newsletter, and an assistive device 
laboratory and loan system. 

2) LOCAL ASSISTIVE TECHNOLOGY 
SPECIALISTS 

A front line or triage resource for Florida’s 
67 school districts is provided through the 
combined efforts of the: 

• Instructional technology staff of the 
18 Florida Diagnostic and Learning 
Resource Centers (FDLRS) and 

• Designated Local Assistive and 
Augmentative Technology 
Specialists (LATS) 

who provide support and resources to 
educational programs of students using 
assistive technology. ATEN sponsors 
ongoing continuing education training and 
consultation to the cadre of 140 LATSs. 

3) DEVICE AVAILABILITY 

A short term device loan library serves as an 
assistive device resource for student 
evaluation and trial use. ATEN manages all 
maintenance, repair and shipping. 



To unlock human potential by providing services 
to support and facilitate quality instruction for 
students with disabilities by providing 
assessment, equipment, and training. 

APPROACH 

Built on a foundation of existing specialized 
support services, the state initiative was 
conceptualized as follows: 

1) THE ASSISTIVE TECHNOLOGY 
EDUCATIONAL NETWORK OF FLORIDA 
ATEN serves as the statewide coordinating unit 
for assisting school districts in provision of * , 
assistive technology serves for Florida students ^ 



4) PERSONAL STUDENT DEVICE LOAN 
A long range goal of the state initiative is 
the institution of a personal student device 
loan program, making any assistive 
technology available to a student free of cost 
when it is necessary for them in order to 
participate in a free and appropriate public 
education. Guidelines have been established 
defining the funding responsibilities of the 
individual schools and the state sponsored 
technology loan pool. 

DISCUSSION 

^ This project has finished its first year. 

/ Q While the personal student device loan 



186 



RESNA’94 • June 17-22, 1994 



Education and Technology 

program is still awaiting funding, the other 
components are fully operational. Community 
awareness workshops have been presented 
throughout the state, in addition to training 
modules on assessment and funding. Voice 
output, written output, computer access, 
light/low tech, vision, hearing, switch 
interface, mounting, and environmental control 
devices are available for short term loan. The 
final projected outcome of this venture is that 
each student will acquire the resources and 
skills necessary to reach his or her greatest 
level of achievement, independence and 
productivity in the community. 

ACKNOWLEDGEMENTS 

Funding for this project was provided by the 
Florida Department of Education, Bureau of 
Exceptional Student Education. 

Catherine H. George 
A.T.E.N. 

434 N. Tampa Avenue 
Orlando, FL 32805 USA 
(407) 849-3504 
FAX 407-849-3518 



ERIC 



RESNA ’94 • June 17-22, 1994 



187 



ACCESS: SPECIAL EDUCATORS ON LINE 



Donna Heiner,Ed.S. R. Hunt Riegel, Ph.D. 

Living and Learning Resource Centre 
Lansing, MI 



ABSTRACT 

The ACCESS Bulletin Board, established in 
January, 1986, links special educators throughout 
the state. Averaging over 30 calls per day, it 
serves as a communication vehicle and training 
device for professionals interested in the 
instruction of students with disabilities. 



BACKGROUND 

Over half of Michigan’s special education 
students reside in the Metropolitan Detroit area, 
which includes Wayne, Oaldand, and Macomb 
counties. In less populated, rural areas of the 
state, students and professionals who work with 
them are distributed over a wide geographical 
area. Personal communication between 
professionals dealing with similar populations is 
limited. In the less urban areas, professionals 
typically serve a diverse population and 
opportunities to learn about advances in new 
instructional methodologies, such as applications 
of assistive technology, are restricted by weather, 
distance, and available Hnances. 

OBJECTIVE 

Project ACCESS was established in 1983 as a 
State Initiated Project by a Michigan Board of 
Education grant under P.L. 94-142. Project 
ACCESS had two primary goals: (1) to provide 
support and assistance to districts in the 
instructional applications of technology for 
special education students; and (2) to provide 
support and assistance to districts in the use of 
technology for the management of special 
education data and planning and monitoring 
requirements. 

Three basic approaches to providing information 
and technical assistance have been used: 



1. A newsletter, distributed to some 18,000 
members of the special education professional 
community in Michigan, was produced 
bimonthly from 1984 to 1993. 

2. A series of training workshops and 
demonstration labs, and a toll-free "hot line” for 
technical help, have been provided continuously 
for the past nine years. 

3. An electronic bulletin board system has been 
operating since 1986 to provide training and 
communications options for the special education 
community. 

APPROACH 

The electronic bulletin board service (BBS) has 
been operating since January of 1986. The board 
has five incoming lines which operate 
simultaneously (one toll free and four regular 
lines). The BBS has received a total of over 
27,000 calls since beginning, and currently 
handles approximately 20 calls per day, seven 
days a we^. 

The BBS uses multi-tasking software ("The 
Chairman") so that sevo:al users can be on-line at 
the same time. 

Present features of the ACCESS BBS include 
such public message boards as: 

AER - Messages dealing with visual 
impairments 

CALENDAR - Upcoming events, workshops, 
etc. 

JOBS - Special education positions available in 
Michigan 

OUTCOMES - Messages from the special 
education Outcomes Training Project 
SIGSPED - An area for communication with the 
special education interest group of MACUL 
TECHBORD - Messages regarding technology 
(help, hints, ideas, etc.) 



188 



■■,:SQ2 

RESNA ’94 • June 17-22, 1994 



ACCESS 



In addition, the BBS pennits establishing private 
conferences (accessible only by members 
approved by their respective group leaders) for 
groups such as: 

MASP - Michigan Association of School 
Psychologists 

MALDE - Michigan Association of Learning 
Disabilities Educators 
MRS - Michigan Rehabilitation Services 
COMPUTE - Coalition of Organizations in 
Michigan to Promote the Use of Technology in 
special Education 

Facilities also exist for uploading and 
downloading of riles, conducting surveys, and 
searching databases for resources and software 
titles. 

Current expansion activities are focused on 
bringing larger databases (such as ABLEDATA) 
on line so those needing information about 
specific assistive technology devices can search 
for product names and producer contacts. 

DISCUSSION 

Three major uses of telecommunication have 
increased dramatically in the paist few years: 

1. Messages to individuals must be sent quickly 
and received when the recipient is ready. 

2. Specifically formatted documents must be 
transmitted so that the receiver can review 
information exactly as it is intended to be seen. 

3. Files must be sent so the receiver can capture 
them for further use on a computer (for word 
processing, data base manipulation and the like). 

A number of methods have been developed to 
accomplish these communications, such as the 
digitized phone service for leaving and receiving 
voice communications and the fax machine for 
sending exact copies of documents. But to rfatp. 
the most readily available method for 
accomplishing all three of the above-mentioned 
uses is the electronic bulletin board system 
(BBS). Such a system allows users to send 
personal messages to each other, post public 
messages for all to read, send documents which 



can be "captured" by other users and printed, and 
send riles to other users which they can c^ture 
for use on their own computers. 

The ACCESS BBS has proven to be a cost- 
effective and viable means of communication 
between professionals who work with special 
education students. Its success is due to several 
factors. The toll-free number encourages use. 
The software itself is user-friendly, with 
extensive prompts in "novice" mode and more 
concise prompting when the user selects the 
"expert" mode. 

At its inception, a small group of individuals 
with the equipment and background to learn how 
to use the system was available. More recently, 
feedback from professionals and observations of 
board use revealed several interesting facts. 
Although more special educators do have 
classroom computers, many teachers do not have 
telephones in their classrooms. The purchase of 
a modem must come from the teacher’s supplies 
budget. Professionals unfamiliar with 
telecommunications require instruction and one- 
on-one support to use equipment and software 
correctly. 

CONCLUSION 

In 1993, the ACCESS Project relocated to a 
State Initiated Project which provides 
information, demonstration, and consultation 
services dealing with assistive technology for 
low-incidence populations. 

The combined staff submitted and was awarded a 
grant to establish MATCH (Michigan Assistive 
Tbchnology ClearingHouse), a component of 
Michigan’s Technology-Related Assistance for 
Individuals with Disabilities Act of 1988. 
MATCH will include a State-wide electronic 
communication dealing with assistive 
technology; the ACCESS BBS will be used to 
facilitate the development of a system designed to 
meet the specific requirements which emerge 
during implementation. The current BBS also 
serves as a training opportunity for those 
wanting to learn how to use this growing method 
for exchanging information. 



Vi 203 

RESNA’94 • June 17-22, 1994 



189 



ACCESS 



ACKNOWLEDGMENTS 
Project ACCESS and the Living and Learning 
Resource Centre are funded through a State 
Initiated Project grant to the Physically Impaired 
Association of Michigan from the Michigan 
State Board of Education. 



Donna Heiner, Ed.S. 

Director, Living and Learning Resource Centre 

601 W. Maple 

Lansing, M 48906 

(517)487-0883 

FAX (517) 487-1605 



er|c 




RESNA’94 • June 17-22, 1994 



TELEPHONE TECHNOLOGY FOR SENSORY INTEGRATION 



MICHAEL E. SWORDS 
SWORDS AND ASSOCIATES 



ABSTRACT 

Using a telephone , can be a formidable challenge 
for those with sensory integration issues. This 
paper discusses difficulties in using the tel^hone 
by a person with sensory integration issues. It 
discusses multisensory techniques and 
technologies to facilitate using a telephone. 



BACKGROUND 



needing remediation are a severe figure-ground 
separation hearing loss (conductive mechanisms are 
within average limits), dyslexic invasions of digits, 
monocular vision, visual nystagmus, eye-hand 
coordination, and finger dextoity deficits. 

These traits combine to make using the telephone a 
challenging task. Problems in using a telephone 
have occurred at home, work, and while traveling. 
Each poses it own unique circumstances and 
requites different remediadoil strategies. 

Three specific at-home telephone tasks need to be 
addressed: 1) obtaining a phone numbo; 2) dialing 
a numba; and 3) hearing through a telephone. 



Sensory integration theory is a concept developed 
by Dr. Jean Ayres.' The concqit is that in some 
individuals a "traffic jam" can occur in the brain. 
A variety of sensory inputs can get "tied up in the 
traffic" of central nervous system processing. 

Multiple body systems are affected. Auditory, 
visual, tactile, vestibular, and proprioceptive 
systems can all be involved. Sense of touch, 
movement, body position, organization of the 
senses, and motor planning can all be affected by 
integrative problems. Learning and many daily 
living tasks can be extremely difficult, if not 
impossible. 

The Ayres Clinic in Torrence California is a rich 
source of information on sensory integration 
issues and thertq)y strategies. 

IMPLICATIONS 



APPROACH 

An eclectic, multidisciplinary tqtproach was used to 
survey methods, equipment, and techniques. 
SimpUcity, low cost, and ease of use of any strategy 
was required. 

DISCUSSIONS AND SOLUTIONS 

1. Obtaining a phone number. 

The local telephone service (Ameritech) provides 
opoator assistance for qualified individuals with 
disabilities. Use of this service eliminates visual and 
motor planning stressors involved in finding a name 
and phone number in a low contrast, fine print, 
phone book. It also eliminates visual focus, 
dyslexic transpositions, and motor planning 
problems when attempting to eitha write down the 
phone number or dial it directly. 



O 

ERIC 



Daily living tasks and stress increase integrative 
problems. Poceptual stress acceloates fatigue. 
Emotional reserves can quickly be depleted. 

Visual, auditory, and motor planning abilities are 
frequently over taxed. Perceptual and integrative 
problems can interfere with competitive use of the 
telephone. 

THE PROBLEM 

This author has a numba of subtle, but highly 
significant, perceptual difficulties. Specific traits \ 



Use of a phone book cannot be eliminated. Perusal 
of the yellow pages is a perceptually demanding 
task. Lighting techniques for low vision are 
excellent strategies for this low contrast reading 
task. 

A lighted magnifier proves to be an invaluable tool 
for telephone book reading. A high intensity task 
light near the phone can assist with reading and also 
with seeing the telephone keys. Perski has 
published other excellent ideas for lighting 
techniques and low vision. ^ 

20 ^ 



RESNA’94 • June 17-22, 1994 



191 



TELEPHONE TECHNOLOGY, 2. 

2. Dialing. 

For frequently called numbers, auto-dialers are 
invaluable. Ones that are built into a phone, as 
well as portable dialers and computer modems all 
help reduce motor planning and visual acuity 
demands. Finger dexterity and hand tremors 
almost become non-issues. 

Manual dialing is a day-in and day-out task 
Larger telq)hone keys increase the visual motor 
accuracy. They also minimize the coordination 
and tremor errors from hitting adjacent keys. 

A phone that displays the numbers on a three- 
eighths inch display helps verify the digits as they 
are keyed. The di^lay also has the advantage of 
retaining a numbo' after it is dialed. In addition 
to instant feedback, it also gives a check in case 
of either a user or system error. This feedback 
helps with self-esteem and confidence. 

A separate in-line speech synthesizer also helps 
with verification and in controlling the speed of 
input. The voice generator will not allow fast 
input of a number. It will only recognize and 
speak a number if the touch tone for a digit is a 
minimum electronic length for decoding. This 
design limit helps reduce motor planning errors by 
forcing a slower pace. If motor coordination is 
working well, inputing the phone number quickly 
makes the synthesizer transparent This is a nice 
feature for others who use the phone. 

Highest levels of successful dials are made by a) 
placing the display so it is directly in the center of 
the visual field, b) turning on the speaker without 
lifting the handset to get a dial tone (eliminating a 
motor planning task), c) placing the written 
number directly under the key pad, d) keying the 
numbers at a pace suitable for processing by the 
voice synthesizer juid watching the numbers 
tqrpear on the display. 

3. Hearing. 

Ambient background noise is an enemy to anyone 
with hearing difficulty. Noise in telephone calls 
can originate at the caller’s location, in the phone 
wiring, or at the receiver’s location. 



Noise at the callers location: 

This is easiest to control. Ambient noise is best 
kept to whisper-quiet levels. When this level is not 
possible, the ear that is not used to listen to the 
phone is plugged. This increases the signal to noise 
ratio and reduces auditory processing demands. 

Plugging of both ears and using amplification to 
overcome the reduction by the occlusion increases 
the signal to noise ratios to very high levels. 
However, this type of boost makes p^eption of the 
low end of a speaker’s dynamic range much more 
difficult Clipping of the speaker’s message is 
frequently a problem with this technique. 

Amplification increases the signal strength. What 
works effectively is a combination handset and 
amplifier. A thumb wheel in the handset provides 
rapid and convenient adjustments that compensate 
for even the softest voices. Occasionally, a speaker 
will have an auditory projection that even with 
mmitnal amplification will cause excessive sound 
pressure levels. Holding the receiver slightly away 
from the ear easily deals with this rate dilemma. 

This type of hand^t also allows the use of long 
cord— a teal plus when tight muscles need to be 
stretched. 

Noise in the telephone wiring: 

Each cotmecdon in a telephone circuit adds 
resistance. Resistances can combine to reduce 
signal strength. All phone lines are directly wired 
to the entrance panel. 

Noise at the receiver’s location: 

People with average hearing can readily hear a 
speaker through typical ambient noises like washing 
dishes or with a radio playing. A simple request to 
eliminate the noise source will generally be 
honored. This reduces auditory fatigue. Similarly 
requesting the other party to use a handset rather 
then a speaker increases signal strength and reduces 
perceptual fatigue. 

These strategies and technologies make using the 
telephone at home far less demanding. One thread 
of perceptual and emotipnal stress is dramatically 
reduced. By using the^e strategies and 
technologies, auditory and emotional reserves are 
Increased. 



206 

RE$NA *94 • Jane 17-22, 1994 



192 



TELEPHONE TECHNOLOGY, 3. 



REFERENCES 

(1) Ayres, A. Jean, Sensory Integration And The 
Child. Western Psychological Services, Los 
Angeles, California. 1979. 

(2) Poski, Thomas, "Illuminating Thoughts: 
Ughting in the Home." Fighting Blindness News, 
pp. 4-6, RP Foundation Fighting Blindness, 
Baltimore, Maryland. Fall. 1993. 

Michael E. Swords 
Swords and Associates 
5505 Lomond Avenue 
Downers Grove, IL <0515 
® 1994(2)« 





RE$NA’94 • June 17.22, 1994 



193 



ASSESSING PREDISPOSITIONS TO TECHNOLOGY USE IN SPECIAL EDUCATION: 
MUSIC EDUCATION MAJORS SCORE WITH THE "SURVEY ON TECHNOLOGY USE* 



Marda J. Scherer & Barbara G. McKee 
Center for Research, Teaching and Learning 
National Technical Institute for the Deaf/Rochester Institute of Technology 

Rochester, NY 



ABSTRACT 



Assistive and educational/instructional technologies 
have expanded the options for individual students. 
For students to avail themselves of the increased 
educational opportunities available to them, it is 
important that they become comfortable with these 
technologies. The Survey of Technology Use is 
discussed as a means of assessing in^viduals’ 
predispositions technology use in general. 

BACKGROUND 



The Survey on Technology Use (SOTU) is a part of 
the Matching Person and Technology (MPT) Model 
and its purpose is to assess predispositions to 
technology use in general. Other assessments in the 
MPT Model are designed to assess the quality of 
match of a person with a specific type of technology 
(assistive, educational, workplace or healthcare). 
The SOTU is a self-report checklist of 29 items in 
four subscales: a) experiences with current 

technologies used (5 items), b) perspectives on new 
technologies (7 items), c) activities (4 items), and d) 
personal/social characteristics (13 items). AJl items 
are in a 3-point semantic differential format (for 
example, depressed, neutral, happy). A companion 
version (Technology Overload Assessment) exists 
for completion by professionals. One such group of 
relevant professionals is music education majors 
who will be teaching in mainstreamed schools. 
Music classes are popular with many students with 
disabilities; the classes are also using more 
computer software packages to teach rhy thm, 
notation, etc. 

RESEARCH QUESTIONS 

1. How predisposed to technology use is a typical 
group of undergraduate music education majors? 

2. How similarly can they rate a student’s general 
predisposition to technology use? 

3. How strong is the SOTU as a measure? 



\ JH 



RESNA ’94 




METHODS AND DATA SOURCES 

Methods used to develop the MPT instruments have 
been described elsewhere. The methods and data 
sources described here are limited to information 
collected to assess the SOTU’s inter-rater reliabiUty 
and stability (test-retest reliability). The SOTU was 
created from the experiences of people who used or 
did not use a technology provided to them and, 
consequently, has content validity . 

Inter-rater Reliability 

Twenty-one music education undergraduate students 
in an educational psychology course at the Eastman 
School of Music/University of Rochester were 
shown a videotape of a student considering using a 
computer and completed the SOTU on that student. 
Item modes were calculated for all items and the 
difference between the mode and the individual 
raters’ response was computed. Since only 
differences from the mode are of interest, the array 
of difference scores was used to calculate the 
"average difference score" or "average deviation 
fi*om the mode." The resulting statistic reveals the 
number of raters who chose the same or similar 
response for a given item. The closer the average 
difference score is to zero, the more consistent were 
the raters’ judgements of the characteristics. 

Stability 

Twenty-two of the music undergraduates completed 
the SOTU on themselves during the first and 14th 
week of the semester (January to May 1993). In 
January, only 22.7% rated themselves as having 
moderate-considerable exposiue to computers; 
another 22.7% had no exposure to computers. 

RESULTS 

Inter-rater Reliability 

Inter-rater agreement for the SOTU items is shown 
in the table on the next page. 



194 



une 17-22, 1994 



Predispositions to Technology Use 



SURVEY ON TECHNOLOGY USE 
Reliability and Validity Data 





Case Inter-Rater Agreement 
(n=21) 


Four Month 
Stability 
(n=22) 






Avg. Deviation 


% Agreement from 


Item 


% Agreement 


from Mode 


Time 1 to Time 2 


Experiences with Current 








Technologies Used 








Socially Enhancing 


81.0 


.19 


72.7 


Satisfying 


90.5 


.10 


90.9 


Enhanced Creativity 


85.7 


.14 


72.7 


Encouraging 


50.0 


.50 


81.8 


Improved Self Esteem 


90.5 


.19 


77.3 


Perspectives on New Technologies 








Exposed 








- As Child 


84.2 


.21 


68.2 


- In Education 


76.2 


.24 


63.6 


- Recently 


45.0 


.55 


68.2 


Challenged 


85.7 


.19 


50.0 


Cognitive Approach: Logical 


71.4 


.33 


57.1 


- Comfortable with 


100.0 


.00 


68.2 


Equipment 








- Use of equipment has 


100.0 


.00 


50.0 


been Reinforced 








Activities 








Active (Sports/Walking) 


95.2 


.05 


54.5 


Hobbies (Satisfying) 


47.6 


.67 


68.2 


Group 


66.7 


.43 


59.1 


Seeks fresh, new activities 


95.2 


.05 


81.8 


Personal/Social Characteristics 








Composed/Calm 


55.0 


.45 


63.6 


Happy 


60.0 


.40 


77.3 


Tolerant 


76.2 


.29 


63.6 


Positive Outlook 


42.9 


.81 


68.2 


Expressive/Outgoing 


61.9 


.52 


63.6 


Patient 


47.6 


.52 


72.7 


Motivated 


50.0 


.50 


68.2 


Persevering 


78.9 


.21 


59.1 


Socially Active 








- Family/Spouse 


50.0 


.50 


90.9 


- Non-Family 


81.0 


.29 


45.0 


Good Sense of Well-Being 


71.4 


.48 


77.3 


Independent 








- Physically 


66.7 


.48 


81.8 


- Emotionally 


70.0 


.30 


54.5 



tm 



COPY AVAILABLE BESNA ’94 • June 17-22, 1994 



195 



Predisposidons to Technology Use 



The first column gives the percentage of raters who 
gave exactly the same response to a given item on 
the first and second administradon of the SOTU. 
The second column is the average deviadon stadsdc 
described earlier. 

Items related to the technology itself and its use 
received the highest consistency in ratings. Items 
concerned with user psychosocial characterisdcs 
received slighdy less agreement scores. 

Stability 

The SOTU seemed to be reasonably stable over the 
four-month period for this group of students. The 
most stability was seen with items related to 
"Experiences with Current Technologies Used;" less 
with those items asking about "Perspecdves on New 
Technologies Used." From Time 1 to Time 2, eight 
students reported being exposed to a new 
technology: 75% of those eight changed their ratings 
for "challenged" and "cognidve approach;" 50% 
changed their rating in "use of equipment has been 
reinforced." Change occurred over time in all 
students’ ratings of their group acdvides and non- 
family social acdvides. Such consistency, the fact 
that these two items both changed, is posidve. 

CONCLUSIONS 

Based on the data from this group of students, the 
SOTU seems to be a reasonably reliable instrument 
with adequate inter-rater and test-retest reliability. 
When asked to complete the SOTU for a 
videotaped student, raters displayed an 80% or 
higher agreement on more than 2/3 of the 
individual items. While not stricdy speaking a 
highly computer literate group, the music educadon 
undergraduates were able to rate a potendal 
student’s predisposidons to technology use quite 
consistendy. When asked to complete the SOTU 
on themselves, test-retest stability was also reladvely 
high. 



Marcia J. Scherer, Ph.D. 

Rochester Insdtute of Technology & 
University of Rochester 
52 Lomb Memorial Drive 
Rochester, New York 14623 
716-475-6598, FAX =716/475-6500 
Internet = mjserd@ritvax.isc.rit.edu 




196 



RESNA ’94 • June 



EXPLORING LEGAL RAMIFICATIONS OF THE APPROPRIATE APPLICATION OF 
ASSISTIVE TECHNOLOGY IN THE INDIVIDUALIZED EDUCATION PROGRAM 



Martell Menlove 
Utah State University 
Utah Assistive Technology Program 



ABSTRACT 

Assistive technology (AT) is redefining educational 
opportunities for childrra served in special education 
programs. AT is being designed, developed and 
implemented for childrwi with an increasingly wider 
range of cognitive and physical abilities and 
disabilities. In the classroom, AT is enabling 
students with disabilities more indepmlence, self 
confldence, production, and integration into the 
mainstream of society (Lahm and Elting, 1989; 
Wilds, 1989). However, most Individualized 
Education Programs (lEPs) for students with 
disabilities are conducted in such a matmer as to 
never address the need a student may have for 
assistive technology. This lack of appropriate 
applications of technology has serious legal 
ramiflcation for those involved in developing and 
implementing lEPs. 

BACKGROUND 

It has been determined that consideration of a child's 
need for assistive technology must occur on a case- 
by-case basis in connection with the development of 
a child's lEP (Schrag 1990). The Code of Federal 
Regulations (34 CFR 300.346) states that a child's 
lEP must include "a statement of the specific special 
education and related services to be provided to the 
child". The above noted Schrag letter states that AT 
qualifies both as a special education service and/or a 
related service. The CFR (34 CFR 300.346) also 
states that the lEP must include "the extent to 
which the child will be able to participate in regular 
educational programs." The appropriate application 
of AT will likely be used in legally defining, and 
possibly altering what has historically been 
determined as, the least restrictive environment 
(LRE). 

Assistive technology is destined to play an 
increasingly important role in the education of 
students with disabilities. The most recent 
amendments to P.L. 94-142 (IDEA) include 
definitions of an assistive technology device and 
assistive technology services. These definitions are 
taken from the Technology Related Assistance to 
Individuals with Disabilities Act of 1988 (P.L. 100- 



407). 

Recent advances in technology and rehabilitation 
engineering have led to a dramatic surge of interest in 
the use of assistive devices, and in the commercial 
availability of equipment and devices designed fo'- 
persons with disabilities. Assistive technology 
includes adapted toys, alternative input and output 
devices for computers, eating systems, powered 
mobility devices, augmentative communication 
devices, special switches, etc. TTiousands of 
cottunerci^y available or adapted devices and 
solutions that improve a student's ability to study, 
learn, compete, work, and interact with family and 
friends, are now available. Access to technology 
promotes efficiency, increases accuracy and often acts 
as a sensory stimulation. However, for those with 
disabilities, it is a vehicle by which many obstacles 
can be circumvented and disabilities overcome 
(Fifield, 1990). In particular, access to technology 
advances holds great promise for enriching educational 
epportunities and effecting the lives of students with 
disabilities (Gradel, 1990)(Barker, 1990). 

OBJECTIVE 

The objective of this paper is to fully explore the 
current legal opinions and documents that have 
reference to the application of technology in 
developing appropriate lEPs for students with 
disabilities. A thorough examination is intended to 
result in greater understanding for parents, special 
educators, local education agencies and other interested 
parties as to what their role and responsibilities are in 
the lEP process. 

APPROACH 

Two major legal barriers to the appropriate 
application of AT for students with disabilities are; 1) 
the lack of information teachers, parents and student 
advocates have about the legal requirements 
concerning the application of AT in the provision of 
special education services; and, 2) inappropriate and 
possibly illegal lEP practices as they relate to 
assessment and meeting the individual needs of 
students. It is not possible at this point in time to 



211 

RESNA ’94 • June 17-22, 1994 



197 



ffiP LEGAL ISSUES 



difTercntiate the information pertaining to these two 
issues because statements and procedures address 
only the provision of special education services. 

In addition to those references presented previously 
in this paper, assistive technology is referenced in 
IDEA rules and regulations as follows: Section 
300.308 Assistive Technology. Each public agency 
shall ensure that assistive technology devices and 
assistive technology services, or both, as those 
terms are defined in §§ 300.5-300.6 are made 
available to a child with a disability if required as a 
part of the child’s - 

(a) Special education; 

(b) Related services; or 

(c) Supplementary aids and services. 

The student assessment/valuation process, as 
outlined in P.L. 94-142, Section 300.532, states 
that ”... the child is assessed in all areas related 
to the suspected disability, including, where 
appropriate, health, vision, hearing, social and 
emotional status, general intelligence, academic 
performance, communication status, and motor 
abilities.” In addition to the traditional assessment 
and evaluation of students, the use of technology to 
facilitate maximum student potential needs to be 
considered (Bragman, 1987). Technological 
considerations augment the traditional evaluation by 
providing information about the student’s ability to 
access and use technology. 

In addition to legislative references to assistive 
technology found in IDEA and its accompanying 
rules and regulations, 3 policy letters have been 
issued for the Office of Special Education and 
Rehabilitative Services responding to inquires 
associated with the provision of assistive 
technology as part of special education service 
delivery. The first letter, August 10, 1990, is in 
response to an inquiry as to whether school can 
presumptively deny assistive technology to students 
and if assistive technology should be considered on a 
case-by-case basis in the development of a child's 
lEP. In reply Dr. Schrag states, ”it is 
impermissible under EHA-B for public agencies 'to 
presumptively deny assistive technology’ to a child 
with handicaps before determination is made as to 
whether such technology is and element of a free and 
appropriate public education (FARE) for that child. 
Thus, consideration of a child’s need for assistive 
technology must occur on a case by case basis in 
connection with development of the child's 
individualized education program (lEP). 

>• 

RESNA’94 • 



A second letter for Dr. Schrag, November 27, 199 1 , 
address several additional question. The first question, 
and the one the pertains directly to assistive 
technology, is whether or not a child may access 
assistive technology for homework, reading books, 
and other assignments away from school. This 
determination is made by the EEP team. "If the EEP 
team determines that a particular assistive technology 
item is required for home use in order for a particular 
child to be provided FARE, the technology must be 
provided to implement the EEP.” 

A third OSERS policy letter dated November 19, 

1993 is in response to the following question: "If a 
students needs a hearing aid (assistive device) is the 
school district responsible for purchasing the device 
under the new IDEA if the device is put on the EEP?" 
The policy letter indicates that historically personal 
devices such as hearing aides were not covered. 
"However, this policy does not apply to a situation 
where a public agency determines that a child with a 
disability requires a hearing aid in order to receive a 
free appropriate public education (FARE), and the 
child's individualized education program (EEP) 
specifies that the child needs a hearing aid.” The letter 
continues with a discussion of assistie technology and 
it can be inferred that this policy applies to other 
assistive devices. Again, it the participants at the EEP 
meeting that determine the what the child requires in 
order to receive a FARE. Accordingly, the public 
agency must make an assistive technology device or 
service available if such device is necessary for a child 
to receive a FARE 

DISCUSSION 

1. Any participant in a EEP has the right to request 
that the EEP team consider the child’s need for 
assistive technology as the EEP is being developed. 

2. If the EEP team determines that the child may need 
assistive technology a part of the sp)ecial education 
delivery, the LEA is responsible to provide the 
technology, including an appropriate assessment to 
determine what the need may be. 

3. Assistive technology can be a form of 
supplementary aid or service utilized to facilitate a 
child's education in a regular educational environment 

4. Assistive technology may be used outside the 
school program if such use is written into the EEP. 

5. School district are generally responsible for the 
provision of all assistive technology that the EEP 
detOTnines is necessary for a child to receive a FARE. 



June 17-22, 1994 



198 



ffiP LEGAL ISSUES 



Recent legislation and administrative mandates have 
significantly impacted the access, societal 
acceptance, and use of assistive technology in 
teaching students with disabilities. New laws, acts, 
and administrative opinions have not only created an 
atmosphere promoting the use of assistive 
technology in education but actually mandate its 
application. Assistive technology is expanding the 
spectrum of least restrictive environments (LRE), 
providing additional means for local education 
agencies (LEA) to provide a better free appropriate 
public education 

(FARE) for students with disabilities, and should be 
an integral part of the individualized education 
program (I^) process. Bergman (1987) states that 
"it is the responsibility of educators to see that 
advanced technology is used to maximize student 
potential and allow the [handicapped] student full 
access to society." 

The application of assistive technology for students 
with disabilities is an emerging topic. At a recent 
Due Process Hearing Officer training attend by the 
author. Art Cemosia, Esq., predicted assistive 
technology to be one of two major areas for 
litigation in special education in the near future. 

This seminar will prepare teachers and parents to 
appropriately apply technology in the lEP process 
and hopefully avoid litigation and thwart Mr. 
Cemosia’s prediction. 

REFERENCES 

Barker, K.L. (1990). Private practice: A successful 
approach to assistive technology service 
delivery. OSERS News in Print, 3(2). 
Washington, DC: Office of Special 
Education and Rehabilitation Services. 

Bragman, R. (1987). Integrating technology into a 
student's lEP. Rural Special Education 
Quarterly, 8(2) 34-38. 

Cramer, S.F. (1992). Assistive technology training 
for special educators. Technology and 
Disability, 1(3), 1-5. 

EDLAW, Incorporated. (1989). Technology Related 
Assistance for Individuals with Disabilities 
Act Alexandria, VA: Author. 



Fifield, M.G. (1990, October). Defining issues within 
the context of assistive technology training. Paper 
presented at the National Outreach training Directors 
1990 Annual Meeting of the American Association of 
University Affiliated Programs, Madison, WI. 

Gradel, K. (1990). Translation of a customer service 
culture into practice in assistive technology 
and employment services. Washington, DC: 
United Cerebral Palsy Associations, Inc. 

Lahm, E. & Elting, S. (1989). Technology: 

Becoming an informed consumer. NICHCY 
News Digest. (13). Washington, DC: 
National Information Center for Children and 
Youth with Handicaps. 

Wilds, MJL. (1989). Effective use of technology with 
young children. NICHCY News Digest. 

(13). Washington, DC: National Information 
Center for Children and Youth with 
Handicaps. 

ACKNOWLEDGEMENT 

Partial funding for this paper was provided under 
Grant #H22A90051 from the National Institute on 
£>isability and Rehabilitation Research and the United 
States Department of Education. The positions and 
opinions expressed in this paper are not necessarily 
those of the National Institute on Disability and 
Rehabilitation Research. 



Martell Menlove 

Utah Assistive Technology Program 
UMC 6855 
Utah State University 
Logan, UT 84322-6855 



213 



RESNA ’94 • June 17-22, 1994 



199 



STUDENTS WHO USE ASSISTIVE TECHNOLOGY: SCIENCE GETS AHEAD OF THE LAW 



Sharan E. Brown, J.D., Ed.D 
Assistive Technology Resource Center 
University of Washington 
Seattle, Washington 



ABSTRACT 

Students who use assistive technology to 
sustain life raise complex clinical/legal 
issues in addition to traditional 
educational/legal ones. The legal rights of 
all students with disabilities in the 
educational setting — including those 
dependent on assistive technology to 
sustain life — have been reviewed in 
numerous other forums. In the following 
article, the author discusses "do not 
resuscitate" (DNR) orders in educational 
settings — a clinical/legal issue which in the 
past was confined to the hospital setting 
but is now also raised in the schools. 
Educational practitioners are facing, with 
increasing frequency, situations for which 
they have received little or no professional 
training or for which there have been 
vague guidelines. DNR orders are 
governed primarily by state health-related 
laws and the author will supplement the 
discussion below with relevant Washington 
State law. 



BACKGROUND 

An observant visitor to any public school classroom 
in this country will be struck by the diversity of the 
student body. Not only do most public schools 
reflect the racial, cultural, and socio-economic 
heterogeneity of the larger society, they also include 
increasing numbers of students with disabilities. 
After two decades of federally mandated 
nondiscrimination towards students with 
disabilities, state and local education agencies are 
becoming accustomed to serving students with 
disabilities in public education settings. However, 
just as the characteristics of the overall student body 
are changing, the population of students with 
disabilities looks different than it did twenty or even 
Eve years ago. Rem^kable medical advances, 
combined with impressive engineering technology, 
have introduced new opportunities for students with 
disabilities to participate in community activities. 



These medical and technological changes, in 
conjunction with the legal mandates to provide 
educational services to children with disabilities, 
have resulted in increasing numbers of children in 
public schools who require assistive technology to 
sustain life. 

In the past, these children simply were not in the 
regular education setting, riding public 
transportation, or interacting with others outside the 
institutional setting. Educators were not approached 
by parents or guardians requesting that they honor a 
students DNR order. The world has changed. As a 
result of the changing school population, educators 
may be asked to honor a DNR order. 

This issue has received minimal judicial or 
administrative attention, but has become a topic of 
intense interest and concern among school staff 
throughout the country. Science has not waited for 
public policy to “catch-up” with the rapid 
technological changes. The public is increasingly 
being challenged by difficult questions concerning 
the right to live or die — and who decides. There is 
clearly no public consensus on any of the variations 
of the right to life issue— as exemplified by the on- 
going abortion discussion. The ethical and religious 
dimensions of the questions will undoubtedly 
continue to be debated for years. But meanwhile, 
the law governing who and how students with DNR 
orders should be served has not kept pace. As 
science has gotten ahead of the law, educators are 
left without clear expectations of their roles and 
responsibilities. 

DISCUSSION 

The well established rule under negligence law is 
that educational staff are responsible for ensuring 
the safety of students in the public school setting 
(1). Teachers and other staff have an affirmative 
duty to aid injured or ill students and act in a 
reasonable and prudent manner to obtain immediate 
care for those who need additional assistance — i.e., 
call parents or emergency medical personnel (911) 
and/or provide transport to a medical facility. The 
case law is replete with examples of teachers and 
school districts found liable for the injuries to a 
student due to acts or failures to act. 



ERJC 200 



RESNA ’94 






’J2U 

June 17-22, 1994 



Students Who Use AT 



Although there are no cases specifically on point, 
one can make the argument that teachers and staff 
working with students who use assistive technology 
may be held to a higher standard of care then would 
be expected of teachers of other students. Teachers 
of children who use assistive technology to sustain 
life are clearly on notice that such students may 
have equipment malfunctions or emergency medical 
needs. Therefore, the “reasonable” person standard 
in negligence law would suggest that the teacher of 
students who use life sustaining technology must be 
prepared for foreseeable problems related to that 
technology. Regardless of the health status of the 
students, the general rule is simple — i.e., all medical 
emergencies occurring at school require the full 
response of staff. 

Applying this general rule becomes difficult when 
the parents of a child have a “do not resuscitate” 
(DNR) order — or some variation of an advance 
directive to a physician to withhold or withdraw life 
sustaining treatment — and ask that the school staff 
honor that request if their child needs emergency 
intervention in school. A school board in Lewiston, 
Maine has recently grappled with this issue when 
the mother of a 12 year-old student with severe 
disabilities asked that her daughter not be 
resuscitated if her heart stopped and presented the 
school staff with a DNR order (2). Originally the 
school board had voted 6-2 to ignore the mother's 
request. However, based on the testimony of the 
girl’s physician, had reversed itself 6-3. Following 
intense protests from teachers and others, the board 
reexamined their decision again. After a hearing 
that lasted until 3 a.m., and included “a procession 
of speakers” offering legal and ethical advice, the 
Lewiston school board reaffirmed its decision to 
honor the DNR order. This decision initiated a 
complaint to federal officials and resulted in another 
reversal. Under the new policy effective February 
1, 1994, teachers are no longer required to follow 
the DNR order and must develop an “individually” 
designed resuscitation plan to be followed in the 
case of emergency (3). 

Many teachers of special needs children have close 
relationships with the whole family and may want to 
be supportive of the parents' request. Other teachers 
may have ethical or religious conflicts with DNR 
orders in general or for a particular child. In 
addition to the difficult personal dilenunas resulting 
from such a request, there are legal ramifications in 
accepting responsibility to honor a DNR order. 
Each state has granted authority to particular 
professions to honor DNR orders. School staff in 
most states do not legally have the authority to 
comply with treatment limiting directives and could 
face serious liability if they do not respond 
appropriately to a child’s emergency needs. 



Response of Emergency Medical Personnel; 
Washington State Law 

Under Washington State law, a physician, health 
care provider acting under the direction of a 
physician, or health facility (4) can honor a written 
directive from an adult or “an authorized 
representative who validly holds the person’s 
durable power of attorney for health care" (5). The 
directive can instruct the individual’s physician to 
“withhold or withdraw life-sustaining treatment in 
the event of a terminal condition or permanent 
unconscious condition” (6). The first limitation to 
note in this statute — commonly known is the 
“Natural Death Act” — is that although an individual 
can make a directive at any time, a physician is 
authorized to act upon it only when one or both of 
the conditions exist — i.e., terminal illness or a 
permanent unconscious condition. 

The second important limitation is that this statute 
does not cover emergency medical personnel who 
respond to an emergency in the community. 
However, another section of Washington law does 
address the situation by stating that the “department 
of health shall adopt guidelines and protocols for 
how emergency medical personnel shall respond 
when summoned to the site of an injury or illness 
for the treatment of a person who has signed a 
written directive or durable power of attorney 
requesting that he or she not receive futile 
emergency medical treatment” (7). 

There were no written guidelines or protocols until 
recently. Emergency medical personnel were 
directed by the Washington State Department of 
Health to provide full response to everyone 
regardless of the existence of an advance directive. 
The individual would then be transported to a 
medical facility where the request could be 
evaluated. 

Interim written guidelines from the Department of 
Health became effective November 1, 1993. The.ce 
guidelines were developed by a working committee 
of approximately 25 people representing the 
interests of the religious community, emergency 
medical personnel, physicians, and the aging 
population. Under the guidelines, emergency 
medical personnel are authorized to honor a request 
for no assistance, if supported by either a bracelet 
obtained from the Department of Health or a 
specific Department of Health form. Any other 
written or verbal directive will not be recognized by 
emergency personnel — in other words they must 
continue to provide full assistance. The Department 
of Health will issue either the bracelet or form to 
individuals with written advance directives that 
fulfill the statutory obligations of the Natural Death 



’ ' •' 21S 

RESNA’94 • June 17-22, 1994 



201 



Students Who Use AT 



Act as described above— i.e., those who are 
terminally ill or in a permanent unconscious state. 
Those with living wills or other written requests will 
not receive the department bracelet or form. These 
limitations of coverage to only these two conditions, 
have resulted in efforts to expand the protocols to 
cover individuals with Alzheimer’s Disease and 
AIDS — which explains why they are called interim 
guidelines. 

Children are not covered under these interim 
guidelines. Therefore, in Washington State, 
emergency medical personnel responding to a 
student in school must provide all assistance and 
transport the child to a medical facility. Emergency 
medical personnel are not currently authorized to 
respond affirmatively to any advance directive 
concerning a child. According to the Department of 
Health, children are not being considered for 
coverage in the final guidelines. 

Policy Co nsiderations 

The decision to obtain a DNR order for a child is 
undoubtedly a difficult one for most families. 
Compounding the moral and ethical dilemma is the 
fact that the law has not kept up with changing 
realities in the community. Some educators and 
parents, among others, may feel that because 
children with DNR orders are now in public 
schools, the advance directive should “follow” 
them. However, in most states, educators do not 
have the legal authorization to respond to a DNR 
order. And, in many states, emergency personnel 
do not have legislative authority either. 

Sufficient numbers of educators and/or parents may 
believe that the laws governing DNR orders in 
educational settings are archaic and infringe on the 
rights of children and/or parents, and advocate for 
change. However, even assuming that there is some 
consensus that DNR orders should follow children 
into the community, how do we safeguard the rights 
of all involved and who should be allowed to honor 
the codes? One option would be for the applicable 
state legislature to “change the law” and introduce a 
bill that grants authority to educators and 
emergency personnel to honor DNR orders with 
whatever restrictions or caveats are deemed 
appropriate. Another option is to clearly give the 
responsibility to the 911 personnel, but continue to 
preclude educators from following the DNR order. 
A third option, and one adopted in Washington, is to 
preclude both emergency personnel and educators 
from following DNR orders for children in a school 
setting. 

As stated earlier, school districts and educators 
confront unique clinical/legal issues when serving 



students who use assistive technology to sustain 
life. As DNR orders become more common, school 
districts will be faced with the necessity of deciding 
how they will deal with parental requests to honor 
them. Should educators honor DNR orders? 
School staff — unlike medical practitioners — have 
not generally been prepared for dealing with and 
accepting responsibility to make choices such as 
this when it involves their students. Fortunately, 
there have been no court cases specifically 
addressing school district or individual educator 
liability in this area. However, the author believes 
this situation will change as the number of students 
who use assistive technology increases unless there 
is open constructive debate regarding the issue. The 
debate will undoubtedly be difficult for all 
involved — regardless of one's personal perspective. 
However, DNR orders in the schools will not go 
away and the issue is too important to ignore. 

ACKNOWLEDGMENTS 

Support for this article was provided in part by the 
National Institute on Disability and Rehabilitation 
Research (grant number H224A30006) funded 
through the Washington Department of Vocational 
Rehabilitation to the University of Washington 
Child Development and Mental Retardation Center 
University Affiliated Program. 

Sharan E . Brown, J.D., Ed.D 
Research Assistant Professor 
College of Education 

Assistive Technology Resource Center WJ-10 

University of Washington 

Seattle, Washington 98195 

206-685-4010 

sbrown @ u. washington.edu 

REFERENCES 

1 K. Alexander & M.D. Alexander, 
AMERICAN PUBLIC SCHOOL LAW, 3rd 
Ed. (1992). 

2 As reported in Board Votes to Honor Request 
Not to Revive Girly New York Times, Nov. 

11. 1993, at B8, col. 1. 

3 As reported in Teachers Now Allowed to 
Resuscitate a Girly New York Times, Dec. 

14. 1993, at A12, col. 1. 

4 Defined as “a hospital ... or a nursing home 
. . ., a home health agency or hospice agency 
... or a boarding home . . . WASH. REV. 
CODE §70.122.020. 

5 WASH. REV. CODE §70.122.010. Parents 

generally are included in this category. 

6 WASH. REV. CODE §70.122.010. 

7 WASH. REV. CODE §43.70.480. 



216 

RESNA ’94 • June 17-22, 1994 



202 



VOICING DYSLEXIA REMEDUTION 



MICHAEL E. SWORDS 
SWORDS AND ASSOCIATES 



ABSTRACT 

This paper describes an adult consumer’s 
impressions for remediating certain dyslexic traits 
using computer voice input and ouQ>ut methods. 

It presents a background to justify using an 
assistive technology ^proach over other 
strategies. 



BACKGROUND 

For this author, sensory integration issues, visual 
disabilities, dyslexia and "learning disabilities" 
have limited competitive performance in school 
and in the work place. A collection of subtle, 
but significant, central nervous system traits 
combine to form a person with a pattern of mixed 
c^)abilities-gifted in some areas and clinically 
deficient in others. 

West has published a highly informative work 
describing the issues for people with a pattern of 
mixed c^abilities.' Gardener’s work also gives 
insight into the benefits, problems, and issues of 
different forms of native intelligences that are 
traditionally unrecognized.^ 

For this author, the pattern of mixed capabilities 
was unmasked in mid-life. Functionally, the 
clinical deficiencies were partially self-remediated. 
There has been a very high personal cost. 

..hree remediation options seemed plausible: 
retraining, personal assistance, and technology. 

1. Retraining can remediate certain traits, but it is 
a long term, time and cost intensive process. For 
example: in a proper psychoacoustic environment, 
lack of phonetic awareness, and in turn spelling 
problems, could be overcome. Time commitments, 
social and professional obligations, and other 
pressures of {^parent adulthood, make a 
commitment to such training an enormous burden. 



RESNA’94 • 



Retraining also poses certain social and financial 
risks. One is the high potential of a mismatch of 
learning style and teaching methods. Anotho* is the 
hard reality that some neurological traits just cannot 
be fixed. For example, in this case typing and 
keyboarding have been a part of everyday life for 
thirty years. Yet, many times a simple word like 
"time" must be entered with conscious 
deliberateness, slowly and methodically, letter by 
letter, spelling it out loud, to avoid it coming out 
"teim." 

2. Personal assistance is wrought with pitfalls. 

First and foremost, are the dependencies issues. 
Self-reliance cannot be fully developed. Ongoing 
cost is a major concern. Reliable staffing is always 
a problem. Outside services frequently are limited 
and costly. 

3. Assistive technology has many pluses in this 
situation. By using this approach, many of the 
specific issues might be clustered. Despite 
perceptual problems, high level computer skills 
have already been achieved. Extensive computer 
experience, and an intense interest in technology are 
strong motivators and bring personal satisfaction. 

Technology appears to be the most cost effective 
investment of time, resources and finances to 
remediate noncompetitive neurological traits and 
disabilities in this midlife adult-in the shortest time. 



THE PROBLEM 

One of the first functional clusters to be addressed 
by assistive technology is written communication. 
Both skills and rate need vast improvements. To 
address both issues, this evaluation tests direct 
computer voice transcription for word processing 
input and voice output as an editing assist. 

Specific clinical traits are: dysgr^hia, dyslexia, 
total blindness in the left eye, lack of smooth 
pursuit in the right eye, and sensory integration 
problems. Most common dyslexic features include 
letter reversals, dropping of word endings, word 



j ■ 

June 17-22, 1994 



203 



VOICING DYSLEXIA REMEDIATION, 2. 

omissions, spelling problems, and phoneme/ 
grapheme awareness. Structural writing errors are 
plentiful. Clinical testing has revealed extremely 
high aptitudes in creativity and vocabulary 
Sensory integration issues are exacerbated by 
physical and emotional stress. Beyond the clinical 
framework, there are the personality issues of 
complex frustration, negative self-imaging, and 
resentments from previous difficult teaching/ 
learning experiences. Dependency on outside help 
for assistance erodes self-confidence and self- 
esteem, and raises frustration levels. 

IMPLICATIONS 

Our increasingly literate culture demands 
sophisticated writing and reading abilities. In this 
situation, assistive technology has the potential to 
fully utilize the gifts, and make the neurological 
traits-that steal commensurate performance- 
transparenL It could provide competitive formal 
correspondence skills, save time and frustration, 
and improve self-confidence and self-esteem. 

Being able to communicate efficiently in writing 
will enhance business communications abilities, 
and widen social and professional avenues. 

APPROACH 

For the voice input portion of the test, a supplier 
furnished, IBM compatible computer with an Intel 
386 processor operating with a twenty megahertz 
clock was used. This machine was equipp^ with 
a noise canceling, head-set microphone^ voice 
processing board, and appropriate software. 

A first draft of a manuscript was dictated into a 
word-processing file, lliese files were then 
grammar and spell checked. 

For the voice ouQrut portion of the test, the 
document file was taken to another location. 

There, it was read back by another specialty 
computer equipped for high quality voice ouqrut. 

For final editing, auditory, visual, and tactile 
senses were combined. During the voice output 
reading, head-phones were used to increase the 
signal to noise ratio. The text was followed word 
for word using a ball point pen to mark errors and 
enhance concentration. 



DISCUSSIONS 

When the user speaks into a microphone, the 
hardware and software decode spoken utterances,, 
into keystrokes for a word processor. It does not 
have the capabilities to decode continuous streams 
of speech. It must have discrete units of sound to - 
decode. This requires the user to make distinct 
pauses between each wwd. 

A criterion for competitiveness is that the voice 
input rate equal ot exceed the evaluator’s maximum 
speeds for handwriting or keyboarding. Maximum 
handwriting speed (for very simple writing) is 
fifteen words per minute, but the writing is barely 
legible and replete with faults. Key-boarding is so 
full of errors that any output rate is meaningless. 

Good writing is an arduous process for anyone. It 
is difficult to separate the arduousness latent in the 
task with the arduousness of the perceptual 
difficulties-an ongoing issue in all areas of daily 
living. To set a reference standard for comparison, 
this study set a ref^nce point of three hundred to 
four hundred accurate, legible, well organized, 
grammatically correct words per hour for simple, 
non-complex writing. This translates to three to six 
words per minute of final ouqmt. 

Maximum capability was tested by speaking a 
tradrdonal typing test sentence twenty times. The 
test sentence was "Now is the time for all good 
men to come to the aid of their country." This 
method allowed the system to build highly accurate 
models of voice patterns. The system achieved an 
accuracy rate of 97% correct at twenty-five words 
per minute. 

The practical output rate was checked by dintarin g 
the first draft of this document Direct transcription 
of these paragr^hs averaged about ten words per 
minute. The slowness comes from having to retrain 
thought and speech patterns to an entirely different 
process of communication, not hardware or 
software. As with all learning, it can be expected 
that, even with complicated composition efforts 
efficiency will improve with expaience. A detailed 
quantitative comparison with the reference standard 
was not accomplished. 

Composition issues aside, a system limit for the 
speed of voice conversion prohibited any increase in 
input rate beyond twenty-five words per minute. 



218 

RESNA ’94 • June 17-22, 1994 



204 



VOICING DYSLEXIA REMEDIATION, 3. 



Two factors prohibit this. First, the test machine 
was a 386/ 20MHZ. With this clock speed, the 
delay time for verification slowed the process to 
unacceptable levels. Second, the structural nature 
of the utterance changes with speed of speaking. 
When the rate of the spoken utterance stream was 
increased beyond twenty-five words per minute, a 
much higher error rate occurred. With a faster 
clock and a retraining of the voice models, 
conversion rate could keep up. 

Using the keyboard to intermittently supplement 
voice input increased productivity dramatically. 

In most cases, a recognition error on a leading 
phoneme was quickly corrected with just one or 
two keystrokes. Errors in recognition occurred if 
a suffix was slurred, or if the voice dropped at the 
end of a word. Then it is necessary to folly type 
out the word. 

Another factor which affects the conversion 
accuracy is the level and type of ambient noise. 
The system is most accurate when it is trained 
and used with a constant level of background 
noise. The microphone supplied with the system 
is a head-set mounted, noise-canceling type. Even 
though this microphone is designed to minimize 
the effects of ambient noise, it is still receptive to 
background and impulse noises. For example, the 
noise of the computer fan is a constant, and does 
not interfere with voice modeling. However, the 
ringing of a telephone will be picked up, and 
confuse recognition of the utterance, causing an 
error. This, of course limits productivity. For 
maximum conversion accuracy and speed, a pin- 
drop quiet ambience is best. 

Spelling errors were minimal. Only homonyms 
needed to be changed. Grammar checking 
revealed a plethora of opportunities for 
improvements in syntax, grammar, and structure. 

Listening to the reading by the synthetic voice 
revealed an additional four per cent of errors, 
beyond spell checking. 

Conclusions— Despite the limitations, the high and 
low extremes for productivity in this test give a 
first approximation of the ctq>abilities of this 
method. When the software has high accuracey 
voice models, the ouq)ut speed t^proaches the 
fastest, barely legible, error laden, handwriting. 
Key-boarding accuracy is dramatically improved. 



Greater accuracy, legibility, and actual yield 
dramatically increase net ouq)ut Spelling errors, 
inversions, and other dyslexic traits are absent 

Multisensory editing, using voice ouq)ut provides a 
final leveling factor for competitiveness with people 
that do not have cognition or sensory issues. 

This test technique gave the writer his highest 
confidence level ever in his written work. For the 
first time, now there is a way to be competitive in 
this critical realm of daily fimctioning. Now, with 
this technology, it is not as necessary to spend as 
much of life’s energy fighting neurological 
predispositions and struggling to maintain personal 
identity. 

REFERENCES 

(1) West, Hiomas. In the Minds Eye, 

Visual Thinkers, Gifted People With Learning 
Difficulties, Computer Images, and the homes of 
Creativity. Prometheus books. 1991. 

(2) Gardner, Howard. Frames of Mind, The Theory 
of Multiple Intelligences. Basic Books. 1985. 

Michael E. Swords 
Swords and Associates 
5505 Lomond Avenue 
Downers Grove, IL 60515, 

© 1994.(2). 



219 



RESNA’94 • June 17-22, 1994 205 



Speech Evaluation of Habilitation Training of 
Hearing-Impaired Children in Shanghai, China 

Shi-hui Xu, MD, Ya-mei QlU, Qi TaO, MSc 
Shanghai First People’s Hospital, Shanghai, China 



Abstract A method of speech evaluation 
of habilitation training for hearing-impaired 
children is described. In this study, nine 
aspects of speech and audition abilities were 
evaluated. The results of following-up in 88 
evaluated children indicated that this method 
was practical, and it was a good way to 
promote the hearing-speech training task for 
hearing-impaired children in Shanghai, 
China. Several factors which might affect 
the results of evaluation need to be carefully 
dealt with. 

Introduction 

Since 1988, with more attention paid by the authorities, 
the work of hearing and speech habilitation for hearing- 
impaired children has developed rapidly in China. So 
far, we have achieved initial success: about 10-'12% 
profoundly hearing-impaired children, after training at a 
hearing-impaired training course, can enter regular 
primary schools to integrate with norma! children. Our 
experience indicates the importance of speech evaluation 
throughout the training course. Evaluation following 
training is essential to detennine whether a child has 
attained the skills taught during the course.* In this 
study we conducted speech evaluation for hearing- 
impaired children following training in Shanghai. 

In order to facilitate and control the quality of evaluation, 
Shanghai (Re)Habilitation Office has been organizing 
speech evaluation annually since 1990. The candidates 
were selected by teachers at training centers or classes. 
The examination board consists of the experts including 
otologists, audiologists and speech therapists. 



2. Words 

(1) comprehension ability: Twenty toys and pictures, 

which were well known to the children, were put on a 
table. The teacher pronounced the name of an objects, 
e.g. telephone, then the child was encouraged to pick it 
out. Altogether 10 words were pronounced and the 
maximum score was 10 points, i.e. one point for one 
word. 

(2) cognitive ability: The teacher showed an object to 
the child, such as a candle, then encouraged the child to 
tell its name, to see if the child could use correct words 
to express what he or she knows. Altogether 10 objects 
were used for each child with the total marks of 10 
points, i.e. each object contains one point. 

3. Short sentences 

(1) comprehension: The teacher pronounced a short 

sentence, e.g. Qing ni nafei zao xi shou. (Please wash 
your hands with soap.) The child should imitate the 
action of washing hands with soap. Ten sentences were 
pronounced, one and a half points for each sentence, i.e. 
the maximum score was fifteen points. 

(2) expression: The teacher showed a picture and asked 
the child to tell the meaning of it, for example Ba ha zai 
kan shu. (Dad is reading a book.) Ten pictures were 
showed, with one and a half points for one picture. 

4. Questions 

The teacher asked a question, then the child answered it. 
For example: Myia zhu zai na er? (Where do you live?) 
Twenty questions were asked, with two points for each 
question. The maximum mark was forty points. 

5. ReoeatinQ sentences 

The teacher read a long sentence, the child repeated it 
immediately, e,g,.Xiao nii hai ba shou pa Hang zai sheng 
zi shang, (The little girl is hanging out a handkerchief on 
the string.) or Zhuo zi shang you mian hao, niu nai he 
guojiang. (There are bread, milk and jam on the table.) 
Fifteen points for 10 sentences, i.e. one and a half points 
for each sentence. 



Method 

An experienced teacher was appointed as the examiner. 
The hearing-impaired children attended the evaluation 
individually. The board members observed and scored 
during the procedure of evaluation. 

Nine aspects of evaluation were conducted to assess a 
child's abilities of speech and audition. 

1 . Chinese phonetic alphabet 

Chinese phonetic alphabet contains 6 vowels, 28 
diphthongs, and 23 consonants. One Chinese character 
is composed of one vowel, with or without consonant, 
such as e (hungry), hua (flower). It is very important to 
learn Chinese phonetic alphabet for a hearing-impaired 
child in learning pronunciation. During evaluation, the 
teacher shows an alphabet, the child pronounces it. 
Twenty-five alphabets were used for each child with the 
total mark of 10 points. Therefore, each alphabet 
contains 0.4 points. 




6. Telling a story according to pictures 

Four pictures were placed in order according to their 
context. The child was asked to tell a short story. Ten 
points for four pictures. 

7. Discrimination of four tones 

In Chinese language different characters may have the 
same alphabets, such as da and da. The meaning and 
character for da and dd are quite different: da means 
strike or hit, while da means big or great. So, it's 
necessary for a hearing-impaired child to practise the 
four tones. During evaluation, four cards were placed in 
front of the child. The alphabet on each card was the 
same, but the four cards presented different tones, e.g. 
da, dd, da, da. The teacher pronounced the alphabet 
with one of the four tones, the child should point out the 
card which contained the tone pronounced by the 
teacher. 

8. Discrimination of environmental sounds 

Ten environmental soufi^s, including thunderstorm, dog 
barking, cock crowing, baby crying, cough, drum, train, 
ambulance, bicycle, fire cracker were produced by a 




RESNA ’94 • June 17-22, 1994 



206 



Speech Evaluation of Habilitation Training in Shanghai 



tape recorder consecutively. The child should tell what 
sort of sound it was. Half a point for one sound. 

9. Speech perception in noisv background 

Fifteen objects and fifteen pictures were placed in front 
of the child. A word such as "banana" or a short 
sentence such as "The little girl is watching TV" was 
produced from the tape recorder with background noise. 
The child should be able to point out the banana or the 
correct picture. Five points for 10 words, and another 
ten points for ten sentences. 

The full mark for the nine aspects above was 150 points. 

Grading 

With the reference to the "Five Grades Standards of 
Aural Habilitation" of Chinese Association of the 
Disabled^, the mark between 141 and 150 points is 
equivalent to Grade I, mark 121~140 Grade II, 91'*120 
Grade 111, Marks below 90 is regarded as fail, which is 
equivalent to Grade IV or V. 

Results of Evaluation 

Since 1990. we have conducted speech evaluation for 88 
hearing-impaired children, among which 17 children 
achieved Grade I, 15 Grade II, 17 Grade 111, and other 
39 Grade IV or V. Those with Grade I and II were 
enrolled into regular kindergartens or nonnal primary 
schools. Among the 17 with Grade III, 14 have entered 
regular school while the rest, together with those with 
Grade IV or V, have been sent into deaf schools. 

Following-Up 

Following-up was conducted to the children who passed 
the evaluation (Grade I to III) and entered a regular 
school thereafter. Their progress, including scores, 
behavior and communications abilities were recorded. It 
was found that the students achieved Grade I habilitation 
were able to .study and take part in various activities 
actively; and most obtained good or excellent marks. 
The students with Grade II habilitation were able to catch 
up with their nonnal peers and/or achieve good marks. 
Of the 17 students with Grade III, 13 were able to catch 
up with the normal hearing peers, but they need extra 
tuition from their parents; one student was left behind, 
and other three entered directly into deaf schools. Some 
cases are as follows: 

I • Ni Zhll had average hearing loss of 80dB SPL(L) / 
70dB SPL(R). After training course of 18 months, she 
achieved Grade II habilitation, and entered a regular 
primary school. She has achieved good .scores and was 
awarded a "Model of self-reliance" last September. She 
is now studying at the 4th grade of the primary school. 

2. Lu Ye had average hearing loss of 80dB SPL(L) / 
85dB SPL(R). Her score was below 90 during the 
evaluation in 1992 and had further speech training for 
one more year. She achieved Grade II during the second 
trial of evaluation. In September, 1993, she entered a 
regular school and has been able to integrate herself with 
no rmal hea ri ng ch i Id ren . 

3. Tao Tong had average hearing loss of lOOdb SPL 
binaural. Following four-year training, he obtained a 
Grade III habilitation and entered a regular primary 
school. However, he was not able to catch up with his 
peers and left the school four months after the entrance. 

4. Peng Chang had average hearing loss of 90db SPL 
binaural. Following the training of three years, she took 
part in the evaluation and reached Grade III habilitation. 




Since that time, she has been studying in a primary 
school for two years. She could follow the teaching with 
effort, and her father had to tutor her almost every 
evening. 

Discussion 

The results of following-up suggest that the speech 
evaluation method used in this study be practical. It 
demonstrates the general level of speech and audition 
habilitation of hearing-impaired children. Grade I means 
that with the help of hearing-aids, the ability of a 
hearing-impaired child in speech and audition may 
approach that of a normal hearing child, with the 
indication of integration with little difficulties. Grade II 
means that these children would be able to integrate with 
other students, with limited difficulties in the class of 
regular schools. Among those of Grade III, some can 
study in a regular school with more or less difficulties, 
while the rest need further training. Grade IV and V 
mean the ability of communications of the hearing- 
impaired children is poor, further improvement is 
necessary in their training center, or they might be sent 
to deaf schools, according to their age. 

The annual evaluation, within whole city range, is a spur 
to teachers, clinicians, and parents. It can raise the 
initiation in the training tasks. Teachers are always 
trying to have more children they taught to pass the 
evaluation. Parents always eagerly look forward to 
having their child to reach a fairly high level of 
habilitation. To our experience, this is a good way to 
promote our work for deaf children. The same method 
of evaluation were adopted by some hearing and speech 
habilitation centers in order to determine, at the end of 
each semester, whether the hearing-impaired children 
learned the skills which had been set out for the 
semester. On the other hand, the hearing-impaired 
children can get more chance to be familiar with the 
method of evaluation. 

There were some factors that might affect the results of 
evaluation in this study, such as: some of the hearing- 
impaired children might be nervous slightly when 
attending the evaluation because many people were 
watching aside; the scoring was sometimes not precise 
during evaluation, e.g. in "Repeating sentences", if the 
child made a few mistakes, the score he or she get might 
not be exactly the same from each board member. 

References 

1. Daniel ling, and Agnes H. Ling: Aural Habilitation, 
the Foundations of Verbal Learning in Hearing-Impaired 
Children, 

2. Chinese Association of the Disabled & Chinese 

Habilitation Research Center: Standard for Speech 

Habilitation of Hearing-Impaired Children, 1989 



Shi-hui XU, MD 
Head 

Dept . of Otolaryngolory 
Shanghai First People ' s Hospital 
190 N. Suzhou Road 
Shanghai, 200085 
CHINA 




June 17-22, 1994 



RESNA ’94 



207 



INNOVATIVE INTERAGENCY COLLABORATION: PROMOTING THE INCLUSION OF A 
STUDENT USING AUGMENTATIVE COMMUNICATION IN REGULAR EDUCATION 

Judy J. R^iii MS, O'l’K Rick Metheny 

Center for Technology in Education Maryland Technology Assistance Program 

Baltimore. MD Hagerstown, MD 



Abstract 

Children with disabihties in Maryland can now access 
assistive technology SCTvices through diif^rat public 
and private agendes. These include assistive technol- 
ogy teams affiliated with several public school 
systems, private health care provide, the Maryland 
Tedmdogy Assistance Progjnm (MD TAP), and the 
Center for Tedmology in Education (CI^. JWbien 
barriers to indqiendence remain desjnte the availabil- 
ity of assistive tedmology services, there is a need to 
cultivate cnidal connections among the players 
(individual consume and agendes) involved in 
serving duldien with spedal needs. This case 
illustrates a powerful model for interagency collabor- 
ation to support the indusion of a stud^t using 
augmentative communication in a regular classroom. 
The outcomes gained through coUaboratiQn produced 
an action {dan which induded immediate steps tow- 
ard an eqm{}m»it loan, staff training, devdofong 
strategies for integration into the dassroom, and a 
method for measuring {nogress. 

Background 

This case study centm around Andrew, a seven year 
old student with autistic tendendes. Andrew's ]>arents 
have invested a lot of time and effort to understand 
his condition and explore strategies that may hdp Him 
succeed. Whai asked about their ex|)eriences as they 
sought additional professional in|mt and suggestions. 
Mom laughed and made some diservations. "PSrofes- 
sionals assume you [parents] don't understand your 
child and your diild's condition. They're trained with 
the skills and [irotocol to evaluate studmts but have a 
tendency to move parents away from the process." 
When Aiidrew moved into a regular first grade 
dassroom. Mom and the school systm requested 
assistance in determining an effective conununicatitHi 
system for Andrew to use in and out of school. That 
differed from jnevious assessments in that the assum- 
ption was that Andrew could use a communication 
system* The objective was to find an effective system, 
not determine whether or not he could use one. 



Maryland TAP, and three members from the CTE’s 
STAT team (an augmentative communication spedal- 
isl, an occu{)ational therqnst, and a physical thera- 
pist). Each mmber of this collaborative team brought 
unique assets to the table. Mom and Dad brought a 
vision for Andrew as well as an unmatched knowl- 
edge of him as an individual. During his entire sdiool 
career, they will be the one constant team member. 
The LEA knew what educational goals Andrew 
needed to meet and what resources, human and 
finandal, they could utilize to meet Andrew's 
educational needs. CTE brought a breadth and depth 
of ex{)erience unavailable locally. MD TAP provided 
short term equi{>ment loan and ongoing training and 
technical assistance to su{>poit the LEA and family. 

The STAT team evolved from a federally funded 
{)ioject desigi^ to train the trainers in assistive 
tedmology throughout Maryland (Fields, 1991). This 
project devdoped a unique model for traimng school 
based {irofessionals through assessments of individual 
students. Training occurs during the visit as school 
{irofessionals gain hands-on ex{ierieiice with s{iedfic 
tedmologies and assessment strategies. The local 
school team then takes pdmary resjxmsibihty for 
following up on next steps idemtifi^ during the 
asmsmi^t 

Objectives 

As consumers and {irofessioDals search for ways to 
assure jprovision of com{)ctent assistive tedmology 
services, it is essential to build collaboration among 
service providers. Collaboration relies on a team 
qiproach in whidi there is res{iect for each members’ 
expolise and willingness to share the res{X)usibility 
for {iroblm sdving and generating {lotential solu- 
tions. The objective was to build a team com{irising 
key players strategically selected for their ability to 
facilitate diange. In this way it was {xissible to 
generate solutions whidi exceeded ^e contributions 
made by individual expert consultants. 

Approach 



The com{>osition of the team which met induded both 
of his {larents, re{>resentatives from his local educa- 
tion agency (his instructional aide, s{)eech language 
{lathologist, case manager, and the director of s{>ecial 
education), an assistive tedmology s{)ecialist from 



The STAT team relies on a collaborative consultation 
{irocess to facilitate mutual {iroblem solving among 
all the {>artid{)ants: consumers, parents, advocates, 
and local school professionals (Locke & Mirenda, 
1992). After giving background information about 



- '222 

RESNA ’!f4 • June 17-22, 1994 



208 



INNOVATIVE COLLABORATION 



the STAT Team during the introductions, there is 
overt acknowledgment of the expertise contributed 
by parents and local school professionals who work 
with the child on a daily basis. There is also an 
established sequence of events outlined for building 
the process of developing creative solutions. This 
includes: 

( 1) reviewing the reasons for referral, 

(2) distinguishing specific technology strategies to 
explore, 

(3) conducting actual technology trials with the 
child in arena style evaluation format (Locke & 
Mirenda, 1992), 

(4) discussing observations about the child’s 
responses as well as elaborating on strategies that 
seemed to work, and 

(5) developing an action plan that targets the next 
steps to take toward implementation of specific 
assistive technology strategies. 

Examples of action steps indude but are not limited 
to; technology loans for trial use, exploration of 
different strategies, fabrication of light technology, 
searching for funding resources, requesting addition- 
al services from other consultants, coordination of 
staff training, and plans for measuring progress with 
implementation. The local school team takes the 
primary responsibility for inq)lementing the action 
plan which they develop. 

Results 

The assessment process revealed that Andrew used a 
variety of augmentative communication strategies; 

(1) imitating sign language modeled during group 
discussions in class, 

(2) independent use of microcomputers for 
practidng math skills at school and for educational 
and recreational activities at home, 

(3) pointing to or physically manipulating objects 
or people, 

(4) informal eye gaze, 

(5) trial use of voice output communication aid 
with an alphanumeric keyboard for three weeks, and 

(6) trial use of a communication board with letters, 
numbers, “yes”, and “no” choices. 

Both parents and school staff agreed that Andrew 
showed a preference for using devices with speech 
output and microcomputers controlled via mouse or 
standard keyboard. Parents wanted to consider 
Andrew’s need for use of facilitated communication 



techniques to increase his participation in academic 
subjects. All participants expressed interest in find- 
ing strategies to promote Andrew’s ability to initiate 
more independent communication in social interac- 
tions at school and at home. 

The trial phase of the assessment found Andrew 
achieved the most interaction using a voice output 
communication aid (VOCA) in the context of a 
familiar art activity. When use of the VOCA was 
modeled, Andrew demonstrated the ability to imitate 
key pressing. As he became engaged in the activity, 
Andrew began to initiate requests for art materials 
using the picture symbol overlay on the portable 
VOCA. He also started to construct messages by 
combining two symbols in sequence. His speech 
language pathologist commented that trial of a 
similar activity using only a picture communication 
board proved less successful. 

When the team discussed observations about the 
technology trials, there was interest in exploring how 
to integrate a VOCA to enable Andrew to initiate 
communication in everyday activities. The question 
about Andrew’s need for physical assistance to 
promote communication posed a challenge to address. 
By keeping focused on Andrew’s needs, the team 
reached a consensus (Grady, Kovach, Lange, & 
Shannon, 1993). It was evident that Andrew requires 
more intensive assistance when he participates in 
particularly challenging academic activities. This 
enabled school staff and his parents to realize there is 
no one conununication strategy that is effective in 
every situation. 

By conducting a careful analysis of Andrew’s partici- 
pation level in daily activities, the school team could 
determine when and how to incorporate the variety of 
communicative strategies Andrew needs to use: sign 
language, speech ouq>ut from a co mmuni cation 
device, and physical assistance to facilitate communi- 
cation by spelling. 

Through this collaborative process, several notable 
results were achieved. Communication strategies 
were developed as a team that could be implemented 
immediately through the^ short term loan of a VOCA 
through Maryland TAP. After identifying individuals 
(the instructional aide, speech language pathologist, 
case manager, and parents) for whom it was ^propri- 
ate, training and technical assistance in programming 
and using the VOCA were made available. The 
director of special education agreed to authorize 
administrative leave and substitutes to allow a speci- 
fied time for planning and training. The school team 
also designated a date to review progress toward 
act^st^ identified during this process. If the 

June 17-22, 1994 



■nr coFi mmMiE 



RESNA ’94 



9 



209 



INNOVATIVE COLLABORATION 

school team found the VOCA an effective comm- 
unication strategy for Andrew, the case manager 
agreed to pursue funding options. In addition, if 
there was a need for staff to receive training in 
facilitated communication techniques, administrative 
support was available. 

Ratings from fe^back forms given immediately after 
the assessment found that participants valued the 
exchange of ideas during the assessment. All 
participants rated the experience as beneficial for 
Andrew. When asked to compare this collaborative 
model to past experiences. Mom responded that *’We 
were actually invited in like everyone else, we were 
part of the team. This process gave my husband and I 
credibility.” In the longer term, this collaboration set 
the environment for what the Director of Special 
Education, described as "...more ongoing cooperative 
planning between the school based team and 
Andrew’s parents . ”. 

Discussion 

Andrew’s case illustrates how a collaborative 
approach to problem solving can facilitate outcomes 
with extensive impact. Each member of the school 
team accepted responsibilities related to their role in 
meeting Andrew’s educational needs. Including the 
director of special education in this process created 
an environment which allowed staff to develop 
solutions with administrative support for implementa- 
tion. 

The success of this collaborative team effort rests, in 
part, on it’s assumptions about who is important to 
include on the team. Virtually anyone who knows the 
student and/or who can impact the selection, acquisi- 
tion, training with and maintenance of assistive 
technology could be included in a collaborative team 
evaluation. This includes teachers, therapists, family 
(parents and siblings), peers, sitters, tectoology 
specialists, funding representatives, voc. rehab., etc. 

A fluid team looks at the needs of the students and 
then assembles a team to address those specific 
needs. 

The factors most critical to the success of a collabora- 
tive team model evaluation include: 

• Thorough background information 

This is imperative to have any degree of effective- 
ness during the team assessment. This collaborative 
team may not have the privilege of first hand knowl- 
edge of the student so it must depend on the back- 
ground information provided by the local team. 



• Clear goals and objectives 

A ‘general technology’ assessment is a ticket to 
disaster, or at least frustration. The visit will only last 
a short time, so having very clear goals will make a 
quantum contribution to the chances of success. 

• Respect for the local team 

They’re the best source of information about what 
might work, or at least clues to what will be effective 
strategies and will be the team that will make or 
break the assessment. Trust them and work WITH 
them. 

• Brainstorming strat^ies 

Not only will this yield creative strategies, but the 
process of collaborative brainstorming invests all 
participants in implementing the action plan. 

It is critical that a more collaborative model be 
adopted at this time because there is an ever-increas- 
ing need for assistive technology to assist with the 
integration and education of students with disabili- 
ties. The onus of implementing those technology 
solutions needs to be on local and regional teams that 
can be replicated rather than on a few individual 
experts. 



References 

Fields, C.D. (1991, November/December). Promoting 
teamwork in Maryland. TeamRehab Report, 40-41. 

Grady, A.P., Kovach, T, Lange, M.L., & Shaimon, L. 
(1993). Consumer knows best: promoting choice in 
assistive technology. PT - Magazine of Physical 
Therapy, 1(2), 50-57. 

Locke, P.A. & Mirenda, P. (1992). Augmentative and 
alternative communication service delivery in school 
settings: review of the literature. Seminars in Speech 
and Language, 13(2), 85-98. 

Acknowledgments 

The authors would like to thank Mary Lynn and 
William Pegg, Andrew’s parents, Sheree Witt, 

Director of Special Education, Allegany County 
Maryland, the Center for Technology in Education, 
the Maryland Technology Assistance Program, and 
most importantly, Andrew. 

Rick Metheny Judy J. Rein 

MDTAP CTE 

1380 Marshall Street 181 North Bend Rd. 

Baltimore, MD 21229 
(410) 646-3000 



■ % Hagerstown, MD 21740 

(301)791-4626 



RESNA ’94 • June 17-22, 1994 



210 



SIG-07 

Technology Transfer 



APPLYING ASSISTIVE TECHNOLOGY CONCEPTS TO NON-DISABILITY PROBLEMS 



Robert A. Chubon 
University of South Carolina 
Columbia, SC U.S.A. 



ABSTRACT 

Evaluation of the instrumentation layout in a lar|;e 
commercial airplane revealed aspects that 
handicapped pilots and contributed to decreased 
flight efficiency. A potential remedy has been 
derived from disability-focused research. The 
experience indicates that there may be many 
opportunities to transfer assistive technology 
concepts to non-disability applications. 

BACKGROUND 

What do a quadriplegic and a Boeing 747 pilot have 
in common? Answer: They are both handicapped by 
aspects of their environment which can be 
ameliorated by a common solution. During a 
review of the coclqrit instrumentation of the Boeing 
747-400, researchers concluded that features of the 
layout reduced the efficiency and effectiveness of 
pilots. Specifically, it was concluded that the pilots 
are "handicapped" by the location and layout of the 
keyboard in its Multipurpose Control Display Units 
(MCDU), which provide access to the plane's 
computer systems. Because of codq>it design 
restrictions, the MCDUs are located in the center 
console where they can be accessed with only one 
hand by the pilot and co-pilot. Moreover, either the 
right or left hand must be used to key in data 
depending on whether the user is seated at the right 
or left of the console. The configuration is such 
that the users' only recourse is to enter data with a 
single finger. 

PROBLEM 

The slowness of single finger data entry discourages 
computer usage, and therefore, results in sub-optimal 
reliance on the plane's computer-based Flight 
Management System. Because computer managed 
flight is more efficient and reduces flight time and 
fuel consumption, failure to use the system 
contributes to higher operating costs. AdditionaUy, 
when data is being entered, the user's attention is 
drawn away from other frinctions. Thus, detraction 
that occurs during data entry is a safety concern. 



APPROACH 

When the researchers recognized that the pilots were 
handicapped by the arrangement, they initiated a 
project directed at improving the situation. 
Relocation and complete redesign of the computer 
data input system were determined to be unfeasible 
because of cost, safety, regulatory, and other factors. 
Consequently, researchers were driven to e;q)loration 
of other means by which to increase data entry rate 
and ease of computer use. When researchers 
conceptualized the problem as that of a 
handicapping envircmment, they began searching 
disability/handicap data bases for help. Their efforts 
led them to the published work of Chubon and 
Hester (1, 2} regarding the enhancement of standard 
computer keyboards for single finger and typing 
stick typing. Subsequently, the researchers applied 
the concepts to the Boeing MCDU problem. 

Efforts centered around increasing the rate of data 
entry. The Boeing researchers were able to use the 
idea of rearranging the keyboard layout to minimize 
finger travel distance, as well as measures to 
facilitate scanning and locating keys. The Chubon- 
Hester layout developed for standard computer 
keyboards and based on letter usage frequencies in 
the English language could not be utilized directly. 
The alphabetic keys on the MCDUs were arranged 
in six rows containing five keys, aeating a nearly 
square configuration. Moreover, the data to be 
entered consisted of individual letters and nonsense 
syllables representing acronyms, abbreviations, etc. 
As Chubon and Hester indicated, however, custom 
minimal finger travel distance keyboard layouts can 
be derived to suit specific user requirements. The 
Boeing researchers conducted letter usage studies of 
actual pilot MCDU input and determined individual 
letter and bigram frequencies. The frequencies were 
then used as the basis for rearranging the key layout 
on the console, clustering the most frequently used 
letters together. In the Boeing layout, the letter A 
was placed at the center because it had the highest 
frequency. The resultant layout was determined to 
require approximately 25% less finger travel than 
the original alphabetically ordered key arrangement. 
Subsequently, the letters denoting directions. 



^ ' 226 

RESNA’94 • June 17-22, 1994 



212 



Applying Assistive Technology Concepts 



NESW, were relocated around the letter A in map 
positions to facilitate location and orientation. 
Although their relocation resulted in a slight 
increase in finger travel distance, it was posited that 
the increase would be offset by diminished scanning 
time. 

DISCUSSION 

Analyses of potential increase in data entry rate and 
computer use indicated that the costs of converting 
the keyboard layouts to the minimal finger travel 
distance conflguration could be recovered in less 
than two years. However, extensive testing will be 
required to evaluate pilot acceptance and other 
human factors. As Chubon and Hester pointed out, 
mathematically correct keyboard models do not 
assure optimal human performance. Negative 
transfer from use of other key arrangements and 
other human factors often weigh heavily in 
determining usage and effectiveness. 

IMPLICATIONS 

Perhaps the most important outcome of this 
ejq>erience, to date, is that it has revealed the 
potential for applying assistive technology concepts 
to non-disability related problems. This was a 
departure from the usual flow of technology 
developments from non-disability to disability 
applications. Expansion of the market for assistive 
technology beyond the disability realm can 
contribute to increased production and lower cost, 
resulting in wide ranging benetit Additionally, the 
experience underscores the importance of publishing 
disability research. 

REFERENCES 

1. Chubon, R. A., & Hester, M. R. (1988). An 
enhanced standard computer keyboard system for 
single-finger and typing-stick typing. Journal of 
Rehabilitation Research and Development. 25(4), 
17-24. 

2. Chubon, R. A. (1988). An inexpensive off-the- 
shelf approach to increasing computer keyboard 
entry rate for single finger and typing stick typing. 
Proceedings of the International Conference of the 
Association for the Advancement of Rehabilitation 
Technology. RESNA: Washington, D. C. 



Robert A. Chubon, Ph.D., Associate Professor 
University of South Carolina 
Department of Educaticmal Psychology 
Columbia, SC 29208. 

(803) 777-8131 
(803) 777-3045 /fljc 

N230108@UNIVSCVM.CSD.SCAROLINA.EDU 



• .227 

RESNA ’94 • June 17-22, 1994 213 



STUDY OF ACCESSIBLE MICROWAVE OVEN DESIGN 



Frederick S. Barthel, BS 
Rehabilitation Engineering Training Program 
University of Illinois at Urbana-Champaign 




ABSTRACT 

This paper explores accessibility of microwave ovens 
to people with disabilities. Two manufacturer's 
microwave product lines are reviewed, determining 
the functional accessibility to the microwave for 
people with different disabilities. Redesign options to 
improve accessibility are presented. 

BACKGROUND 

The question exists as to how accessible technological 
products are to the more than 40 million American 
consxuners with disabilities. (1) Besides preventing 
the individual with a disability from enjoying modem 
conveniences and performing simple tasks taken for 
granted by the rest of society, denial of access to 
these products is discrimination. Ideally, commercial 
consumer products could be designed from the 
beginning so that anyone could use them, regardless 
of ability. In producing a product accessible out of the 
box, the market for that product is expanded to 
include people with disabilities. This is a concept 
known as universal design. In many cases, relatively 
simple design changes by an informed designer can 
increase accessibility and ease of use for everyone. 
While some research exists concerning universal 
design and application to the disability community, 
little focus has been placed on specific consumer 
products. (2) The microwave oven is chosen here as 
a specific and relevant consumer product to study. 

OBJECTIVE 

In analyzing the accessibility of the microwave oven 
product, it is useful break down and evaluate the 
different functional aspects of microwave usage. The 
functions can then be reviewed for accessibility 
relative to different disability categories. Good design 
features and potential areas of redesign can be 
identified, providing information to manufacturers to 
increase understanding of accessibility issues. In the 
process, consumers are informed of the current 
accessibility status of technology in general. 

METHOD 



representative manufacturers of microwave ovens. 
First, a detailed analysis of specific microwave 
models was performed, concentrating on frmctional 
characteristics. The major task involved in microwave 
usage is heating a food item. All fimctions directly 
involved with achieving the major task are 
categorized as essential functions. Such essential 
functions include: 

• opening/closing of the door 

• insertion/extraction of food item 

• setting power level and cook time 

• starting/stopping the cook cycle 
Additional nonessential functions (those functions 
available but not necessary for completion of the 
major task) include: 

• setting the clock 

• setting a timer 

• using a heating probe 

In addition, the consumer help service was contacted 
for each company to gain additional literature and 
information and also to gain further insight into the 
sensitivity of the company to accessibility issues. 

RESULTS 

While the assorted models of microwaves produced 
differ somewhat in appearance, a "standard” version 
of microwave oven becomes clear after evaluation. 
For instance, the vast majority of available 
microwaves have a flat membrane keypad by which 
to enter commands and control functions such as 
power level and cook time. Primary feedback on these 
functions is generally given by a digital display panel 
while secondary feedback comes through tones 
sounded as buttons are pushed and when cook cycle 
is complete. Of the nearly 60 different models of 
microwaves produced by GE and Whirlpool, all but 
four models are configured this way. The remaining 
models (all compact or sub-compact) have dials or 
function control knobs and analog status displays. In 
addition, door access generally occurs by overcoming 
a friction latch by pulling the door handle or by 
pushing a button to release the latch. 

DISCUSSION 



Whirlpool and General Electric were chosen as 



The term disability covers a broad range of 
conditions. Generally, though, disabilities can be 



214 



RESNA ’94 • June 17-22, 1994 



Microwave Accessibility 



placed in one of four categories: physical disabilities, 
hearing impairments, visual impairments, and 
cognitive disabilities. Discussion of accessibility to the 
microwave is done in terms of accessibility for the 
different disability categories. For sake of simplicity, 
multiple disabilities are not addressed. 

Physical Disabilities (impaired range of motion, 
paralysis/amputation, weakness, control difficulties, 
and impaired tactile discrimination): 

Food Manipulation: It is important to realize that 
some limitations preclude the use of a microwave, 
regardless of the microwave itself, by preventing the 
completion of the essential function of 
insertion/extraction of a food item. For example, 
someone with extreme control limitations may have 
great difficulty placing a cup of water in the 
microwave without spilling. The success of such 
manipulations are dependent on the size, shape, and 
type of the food item. However, of primary concern 
here is the actual accessibility of the microwave rather 
than the food items which must be manipulated. 
Door Opening: A person with weakness, limited use 
of arms and hands, or control difficulties may have a 
lot of trouble grabbing the door handle and pulling it 
open. ADA building guidelines specify that doors 
should require less than 5 Ibf to open, and it is 
reasonable to transfer this guideline to microwaves. 
Most handled models of microwave require between 
7 and 15 Ibf to open - more than desirable. For 
models which require pushing a button to release the 
door, generally less than 4 lbs of force is required, 
but fine motor control is necessary to be able to 
activate the door release button. 

Keypad Button Activation: These buttons tend to be 
small ('^1/2" square) and therefore pose a problem for 
people who lack fine motor control and are unable to 
accurately select a button without inadvertent 
activation of neighboring buttons. Also, there is no 
tactile feedback given to indicate that a button is 
pushed, possibly contributing to selection difficulty. 
Temperature Probe Use: Microwave models which 
have this feature utilizes a 1/4” phone jack/plug as a 
connector. The placement of the jack inside the 
microwave is critical in determining the awkwardness 
of the motion of plugging in the probe, especially for 
an individual in a wheelchair. One G£ model placed 
the jack on the rear part of the side wall, almost a 
foot from the front of the microwave. It would be 
awkward to try to reach all the way into the 
microwave and try to guide a plug into the jack. A 
Whirlpool model more appropriately placed the jack 
as close to the front of the microwave as possible. 



hearing loss): The facet of microwave function which 
most effects an individual with a hearing impairment 
is audio feedback. For most models of microwaves, a 
tone is given when each selection button is pressed. 
Inability to hear this would interfere slightly with 
ability to enter cook time and power - essential 
functions. Also, all microwaves give some sort of 
audio signal when finished with a heating cycle. 
Usually, though, there is also visual feedback coupled 
with the audio feedback in the form of status readings 
on the digital display. Therefore, the only real concern 
for people with a hearing impairment is knowing 
when the cook cycle is finished when they are not in 
close proximity to the microwave, and therefore 
cannot see the display reading. This difficulty would 
not prevent the completion of the major task, and it 
can be concluded that microwaves are quite accessible 
to people with hearing impairments. 

Visual Impairments (low vision and blindness): Of 
major concern for people with visual impairments the 
amount of visual information which is presented in 
relation to the microwave for identification and 
discrimination of functions and status. 

Feedback In terms of feedback, there is no tactile 
feedback which would indicate which button is being 
activated or even where the buttons are located, and 
so individuals with a visual impairment are forced to 
rely on the limited audio feedback described in the 
section above. Also the digital display tends to be 
small and hard to read, so visually impaired 
individuals would have a difficult time utilizing that 
feedback as well. One difference between 
manufacturers is in status indicator labels such as 
"power” or "cook level.” All GE models include these 
in the digital display, and so the words are at most 
1/16” tall. Whirlpool has some models in which these 
labels are printed on the membrane with LED arrows 
pointing to the appropriate words. This allows the 
status labels to be printed larger and be easier to read. 
Labelling: The major problem with labelling is 
difficulty in differentiation between keypad buttons. 
Both companies indicate that braille overlays for the 
keypad are available for their microwaves, but braille 
is not useful for the 70% of visually impaired 
individuals who do not read braille. Also, the GE 
braille labels are stickers with a rather cryptic coding 
system which must be applied by a sighted person (as 
stated on the printed instruction sheet sent with the 
braille labels). The Whirlpool version is a true overlay 
which can be installed by someone who is blind. 
Whirlpool also offers the Use and Care Guide and a 
cookbook in braille, large print or audio tape. 




Hearing Impairments (hearing deficits and total Cognitive 




Disabilities 



RESNA*94 • June 17-22, 1994 



(memory impairments. 



215 



Microwave Accessibility 



language deficits, and difficulty with abstractions): 
Button sequencing memory: The method for setting 
power level, setting cook time, setting the clock, and 
setting the timer are each very similar for each 
membrane keypad microwave. Each function requires 
a sequence of 4-5 key depressions, making the total 
number of depressions for the whole task at least 12 
for the most complicated plan. Most models have 
simpler settings which would allow power and time to 
be selected by a single touch. A sequence of three is 
much simpler to remember than a sequence of twelve. 
Label Abstraction: Some models have pictures or 
symbols which need to be interpreted in order to 
make use of them. For example, a picture of a 
popcorn kernel may set the time and power level as 
appropriate for microwave popcorn. For the person 
making popcorn, this is fine. For the person cooking 
vegetables, however, it may not be simple to think 
that vegetables = popcorn, even though the time and 
cook level may be completely appropriate, and so the 
individual may unnecessarily go through the longer 
sequence of button pushing. There is a high level of 
abstraction here which must be considered. 
Literature: A complex and wordy owner’s guide will 
make it difficult for some individuals to understand. 
In general, the Whirlpool manual contained simple, 
step-by-step instructions which were accompanied by 
pictures. The GE manual tended to have longer 
narratives and was relatively more difficult to follow. 

CONCLUSION 

The following is a list of design features which would 
make the essential functions associated with the use 
of a microwave oven more universally accessible. 
These features will not only aid people with 
disabilities, but also make the microwave oven easier 
and more convenient to use for all people. 

Levered Door Latch: A push button release 
mechanism is recommended that has a large activation 
area and makes use of leverage to decrease the 
necessary applied force. The door should be able to 
be opened with a closed fist. 

Large Button Models: Large buttons would not only 
assist those with visual impairments to see the buttons 
and labels more easily, it would also help those with 
control difficulties to accurately make a selection. 
Availability of a keyguard would also be appropriate. 
Large Display: A large display would improve 
readability of the visual feedback. 

Voice Output Feedback A means for the microwave 
to ’’speak” the buttons as they are pushed and a 
method to "read” the display would serve to give the 
visually impaired audio feedback to aid in their use of 
the microwave. Also, for a person with a language 

2 

RESNA’94 • 



deficit (e.g. a learning disability which hinders them 
from reading numbers correctly), the voice output 
could greatly clarify the status of the microwave and 
make sure that commands are entered correctly. 
Single Touch Settings: Many models already have 
single touch settings which avoid long activation 
sequences and aid in the cook cycle programming. 
Tactile Feedback in Buttons: Allowing the user to feel 
^en a button has been activated will help those who 
cannot use the visual or audio feedback. 

Readable Literature: All literature associated with the 
microwave should be as readable as possible for the 
user. First, it needs to be in a form t^t the user can 
access * braille, large print, and audio-taped versions 
are a very good idea. Also, the actual writing of the 
literature should be clear and concise, with 
instructions simple, warnings obvious, and utilizing 
pictures and diagrams as much as possible. 

Both manufacturers are definitely aware of 
some needs of people with disabilities, but there is 
always room for improvement. With increased 
information and sensitivity concerning people with 
disabilities, these and other companies are on the road 
to providing consumer products which are more 
useable by everyone. 

REFERENCES 

(1) Strauss, M.G. (1992) Promoting Ability with 
Technology . Vol 1, Version 2.1, p. 20. 

(2) Vanderheiden, G.C. and Vanderheiden, K.R., 
(1991). "Accessible Design of Consumer Products." 
Working draft 1.6. 

ACKNOWLEDGEMENTS 

Support for graduate student performing this research 
provided by RSA grant #H129E20001. The author 
gratefully acknowledges Dr. Jon Gunderson for his 
input and assistance. Whirlpool and GE for their 
cooperation, contributions and interest, and 
Champaign Appliances for hosting the on-site survey. 



Frederick S. Baxthel 
Rehabilitation Engineering Program 
University of Illinois at Urbana-Champaign 
1207 S. Oak St. 

Champaign, IL 61820 
(217) 244-8265 
(217)333-1970 (TDD) 

(217)333-0248 (fax) 
e-nml: barthel@uxl.cso.uiuc.edu 




June 17-22, 1994 



216 



Accessibility Evaluation of Cunent Television Design Trends 



Brian T. Fay, BS 

University of Illinois at Urbana-Champaign 
Rehabilitation Engineering Program 



ABSTRACT 

The accessibility of television to people with 
disabilities is becoming more important as 
television provides greater news, information, and 
entertainment services. This paper reviews the 
current accessibility of television sets. How well 
current television controls and displays address the 
requirements of people with physical disabilities, 
cognitive disabilities, visual impairments/blindness, 
and hearing impairments/deafness are considered. 
Possible redesigns are discussed. 

BACKGROUND 

Researchers have recently attempted to formulate 
methods of rating product accessibility for all 
disability types [1,2]. Often this has entailed 
reduction in the number of distinctions between 
different types of disability in an attempt to produce 
a 100% accessible product design criteria [3]. 
While the validity of such imiversal design is still 
debated, these studies provide a good resource for 
determining possible product design changes. 

INTRODUCTION 

The Americans with Disabilities Act (ADA) and the 
growing involvement of people with disabilities in 
the consumer movement have created an 
atmosphere through which citizens are achieving a 
higher quality of life. Television is a very popular 
form of information and entertainment 
dissemination in our society today. It is therefore 
reasonable for people with disabilities to both desire 
and expect the accessibility of television should 
they choose to use it for entertainment, information, 
etc. 

Currently, television sets have many features which 
may greatly enhance out-of-box accessibility if 
certain changes were made in their application. It 
is proposed to review current trends in television set 
design to determine which changes are appropriate. 
The television models reviewed are the Zenith 
and Quasar™ brands and other after market 
products such as remote controls. 

msnissioN 

Accessibility was evaluated for each of four 
disability groups: (1) Physical Disability, (2) 



Cognitive Disability; (3) Visual Impairments/ 
Blindness; (4) Hearing Impairments/Deafness. 

Physical Disabilities: Physical disability lends itself 
more to access difficulties due to the television set's 
controls. 

• Action: Push buttons switches appear to be 

superior to other switches in this case. 
Prehension of pull and rotary switches can be 
difficult. Both television lines used push buttons 
on the television set and the remote control. 

• Location: Placement at a height between 

waist and chest on the front of the set is 
important for people in wheelchairs or people 
with limited range of motion. Both brands placed 
push button controls centered below the screen. 
Placement with sufficient surface area to prevent 
activation of surroimding switches is significant. 
Zenith™ sets had 3/8" x 3/4" push buttons 
successively in a line. Quasar™ sets placed 1/8 
roimd push button controls behind a flip down 

door. This placement made it very difficult to 
access the controls even for able-bodied persons. 

• Activation Force: This should be small enough 
for individuals with low strength to actuate the 
switch, but a delay on reactivation would be 
advantageous to people with linuted control. 
Zenith™ sets had activation forces of 225g 
whereas the Quasar™ sets were lOOg. In 
comparison, lift switches usually require about 
300g implying these forces are not excessive. 
Neither brand had a delay feature on their 
switches. 

• Texture: Switches with dimpled or friction 
surfaces prevent slipping of the actuator, be it a 
finger, mouth stick, etc., off the switch surface. 
Zenith™ switches are smooth without a friction 
surface, whereas Quasar™ controls are roughened, 
but are very small. Changing to a switch with a 
depressed surface may help to keep one s finger, 
mouthsitck, etc. on the switch. Many remote 
controls have friction surfaces which adds to their 
accessibility. 

• Shape: Shape seems to be less an issue if 

push buttons with sufficient surface area are used. 
With pull or rotary switches, surfaces with 
varying diameter and splined sides respectively 
are appropriate. The small size of the switches 



2 / 31 ., 

RESNA ’94 • June 17-22, 1994 



217 



Television Accessibility Evaluation (continued) 

on the Quasar™ sets left them relatively useless. 

• Feedback: Regardless of switch type used 

there should be at least two types of feedback: a 
tactile and audible "click". Zenith™ sets 
provided both modes of feedback, whereas the 
Quasar™ sets provided little tactile and no 
audible feedback. Remote controls with this 
feature have increased accessibility. 

• Labeling: Text, symbols, and color should be 
used to allow discrimination of the power switch 
from other switches. Switches on the set did not 
always display this feature. However, the remote 
controls had red colored |X)wer switches. 

Cognitive Disabilities: Cognitive disabilities lend 
themselves more to difficulties with displays. 

• Information Presented: The amount of 

information placed on the control pad of the TV 
or remote control and resulting screen displays 
can play a large role in confusing not only people 
with cognitive disabilities but also the general 
public. Zenith™ low end remote controls 
presented information in an uncluttered simplified 
fashion, but higher end models of both brands 
came with relatively complicated button layouts. 
Quasar™ screen display of functions such as 
volume, picture, and channel are exceptional. All 
functions are presented on the screen in three 
colors with both text and symbols. Movement 
between main and submenus is easy and intuitive. 

• Presentation Format: Formats which rely 

solely on printed text for the relay of information 
may create a handicap which would not exist if 
the display included abstract or concrete 
symbols. Zenith™ sets used symbols in a limited 
fashion whereas Quasar™, through on screen 
displays, used symbols extensively. 

• Display : Information presentation should flow 
either from left to right or top to bottom. 
Running level bars (ie, volume) should rise (top) 
and fall (bottom) with increases and decreases in 
output. Both brands used horizontal bars to 
display information. Also, headings should be 
prominent, that is highlighted or in a different 
color. Both brands used this feature effectively. 

• Location: The placement of information 

should be easily seen. If on the control pad or 
remote control, the button or switch identifier 
should be obvious. Placing switch labels on the 
switch instead of next to them aids accessibility. 
Also, positioning switches in easy to reach areas 
reduces the possibility of an added frustration. 

• Sound: If a sound occurs during manipulation 
of television functions, the sounds should be 
distinct and at a relatively high or adjustable 

RESNA^94 • 



volume. Neither brand demonstrated this feature. 

• Visual: The size or height of text should 

allow for easy reading, color should be used to 
enhance discrimination between different 
functions. Both brands used this feature well. 

• Tactile: Switches and remote controls should 
exhibit the use of different button shapes, sizes, 
and textures. The use of patterns is another 
technique to enhance accessibility. Neither brand 
used raised letters or numbers on the set or on the 
remote control. Buttons in the shape of their 
function such as an arrow directed up for higher 
volume or channel is more intuitive. 

• Feedback: Buttons should "click" both 

audibly and tactually. On screen running displays 
(ie, volume) may also provide intuitive feedback. 

Visual Impairments/Blindness: A colleague asked 
why I was bothering with this group of people with 
disabilities. Visual Impairments and blindness do 
not end a persons desire to keep up with the news 
or enjoy a drama or comedy. Television 
accessibility to this group is as valid as it is for the 
general public. Obviously, visual displays are of 
less concern here, however this increases the need 
for accessibility to other methods of interaction. 

• Format of Presentation: Most people with 

visual impairments and blindness would have 
little trouble accessing a television if only the 
controls were arranged in an intuitive manner. 
The use of raised numbers or button shapes 
relative to their function may accomplish this 
goal. With continuous use of a remote visually 
impaired persons may become used to the layout. 
Some brands of remote controls use raised 
numbers and letters and shaped volume, channel, 
and VCR buttons, or have certain buttons with 
dimples to serve as landmarks. 

• Visual: The proper sizing of screen text, or its 
adjustability, can determine whether a standard 
feature allows access. Neither of the brands 
allowed for screen text sizing. 

• Control Action: The rotary switch is often 
considered to be the most intuitive switch for 
people in this group. However, if the controls are 
arranged in an intuitive manner with some type of 
landmark identifying the control layout, then the 
controls become useful and push buttons may 
allow access. 

• Texture: Braille is based on outward dimples. 
Possibly dimpled overlays could be offered. 

• Shape: The shape of the control is probably 
the most powerful design change that could be 
made for this group. Controls with raised shapes 
or buttons with shapes indicative of the control 

12 

June 17-22, 1994 



218 



Television Accessibility Evaluation (continued) 



function greatly increase accessibility. 

• Labeling: Labeling need not only be in 

painted text, but could also be raised text. 

• Feedback: Feedback in the form of sound or 
tactile is important since the result of control 
activation on the screen may not be seen clearly 
or at all. Possibly an add-on box could provide 
audible identification of channels. 

Hearing Impairments/Detrfness: Accessibility for 
people with hearing impairments and deafness has 
improved with the mandatory inclusion of a closed 
captioning chip in every television set. However, 
this has opened new issues such as how to best 
display text on the screen. 

• Format of Presentation: The way in which 

captioning is presented should allow viewing of 
other on screen information. Some sets allow the 
user to expand the captioning from the normal 
three lines to the whole screen. However, when 
this feature was activated the picture was covered 
with a black background. 

• Visual: The correct sizing of screen text, the 
rate of text presentation, and the contrast of text 
determine readability. It is most effective to 
maintain good contrast between text and 
background since the rate of speech cannot be 
controlled. 

• Feedback: Tactile feedback upon control 

activation is important for this group. 

RECOMMENDATIONS 

Bringing these design changes for all disability 
groups together into a more accessible television 
can be achieved by defining fo\ir basic functions of 
television operation: Power On/Off, Volume 
Adjustment, Changing Channels, and Picture 
Adjustment. 

Power On/Off The switch should be a push 
button type of sufficient smface area, low activation 
force and a delay on reactivation. A friction or 
depressed surface and tactile and audible feedback 
is likewise advantageous. This switch should be 
labeled by using raised text or symbols and the 
effective use of color. 

V olum e A djustm ent Contro Is should be 
arranged vertically with either function 
representative shapes or raised text or symbols. 
Sufficient space between switches is important. 
Visual feedback should be supplied on the screen 
possibly in a vertical running bar. Closed caption 
should provide good contrast, while also allowing 
view of the screen. Possibly an auxiliary stereo 



headphone jack with adjustable volume could be 
supplied so people with hearing impauments could 
watch along with people with better hearing. 
Braille overlays may add additional accessibility. 

Changing Channels Controls should be as for 
volume adjustment. Visual feedback on the screen 
and possibly voice feedback for people with visual 
impairments or blindness. 

Picture Adjustment Simplified control layout on 
the remote controls can aid the comprehension of 
this function. Screen display with good use of 
color, symbols and text are important. 

Of these four functions, access to power on/off, 
volume adjustment, and changing channels could be 
considered essential functions since they are 
commonly used. Picture adjustment is usually not 
reset often. Many current sets have out-of-the-box 
factory adjusted pictures. As such, picture 
adjustment is less essential as the other functions. 

REFERENCES 

[1] Kanis, H.; "Design for All? The Use of 
Consumer Products by the Physically 
Disabled"; Proc, Human Factors Society\ 32nd 
Annual Mtg, 1988, 416-419. 

[2] Hollerith, R.; "Make Products More Usable 
for Those with Disabilities"; Proc, of the 
Symposium on Human Factors and Industrial 
Design in Consumer Products\ May80; 83-95. 

[3] Ibid., 88-89. 

ACKNOWLEDGEMENTS 

Funding for this paper was provided by the 
Rehabilitation Services Administration (grant 
number H129E20001). 

Manny & Martin's TV & Appliance; Champaign, IL 
for allowing us to review the television sets. 

Brian T. Fay, BS 

University of Illinois at Urbana -Champaign 
Rehabilitation Engineering Program 
Dept, of Rehabilitation Education Services 
1207 S Oak St 
Champaign, IL 61820 USA 
(217) 244-8265 (voice) 

(217) 333-1970 (TDD) 

(217) 333-0248 (fax) 
email: btfay@uxa.cso.uiuc.edu 




June 17-22, 1994 



RESNA ’94 



219 



SUPPORT FOR TECHNOLOGY TRANSFER IN THE 
ONTARIO REHABILITATION TECHNOLOGY CONSORTIUM 



William G. Bennett, Morris (Mickey) Milner 
The Hugh MacMillan Rehabilitation Centre 
Toronto, Ontario, CANADA 



ABSTRACT 

This paper describes the approach that a rehabilitation 
research and development organisation h^ taken to 
fulfil its mandate to "address the technology needs of 
consumers, while building a stronger research base to 
effectively transfer relevant technology to the 
industrial sector." Four factors which are useful for 
the promotion of an organisation’s product 
development culture are delineated. 

BACKGROUND 



Rehabilitation research and development is part 
science but is at least in equal part an art of product 
design. In this latter part consumers' opinions are by 
definition the criteria of acceptability. Since there are 
fewer consumers of rehabilitation products than of 
other commercial products, and since there is less 
competition and hence less choice in the marketplace, 
the Consortium has created formats through which 
informed consumers can contribute and participate in 
the development of assistive devices. 

OBJECTIVES 



In 1991, the Ontario Ministry of Health established 
the Ontario Rehabilitation Technology Consortium 
(ORTC), consisting of a collective of teams 
investigating the design and appropriate use of 
rehabilitation technology (1). Members of the 
Consortium are affiliated with: the Hugh MacMillan 
Rehabilitation Centre, Ontario Institute for Studies in 
Education, the Ottawa Rehabilitation Centre, Queen’s 
University, Sunnybrook Health Science Centre, the 
University of Toronto, the University of Western 
Ontario, the University of Waterloo, and West Park 
Hospital. 

Eight teams are curmetly in place, with the following 
areas of interest: 

• Commmunications 

• Hearing 

• Mobility 

• Prosthetics and Orthotics 

• Psycho-social Evaluation 

• Respiration 

• Seating 

• Vision 

A diverse group of Ontarians is involved in this 
challenge to bring together scientists, consumers of 
assistive devices, industrial representatives, and health 
care providers from across the province in a program 
of research and development activities designed to 
meet the objective of bringing new assistive device 
technologies to consumers as rapidly as possible. The 
Hugh MacMiUan Rehabilitation Centre (HMRC) has 
facilitated the coUaboration, and serves as 
headquarters of the ORTC. 



The Consortium's activities focus on assistive devices, 
which have been defined as ''devices which assist 
individuals to compensate for physical deficits in 
school, work and community living." The Consortium 
is mandated both to research assistive devices and, 
through technological advancement, to contribute to 
the economic development of the Province. 

Each of seven teams is currently involved in two 
major project activities, for which the emergence of 
products is intended. A Technology Transfer Unit 
interacts with the teams to facilitate technology 
transfer, and growth and development of Ontario 
industry in particular. A Psychosocial Evaluation 
Team, working with all of the other teams, assists in 
evaluation, encouraging and ensuring appropriate 
consumer involvement. 

METHODS 

To meet ambitious technology transfer objectives, a 
rehabilitation research group needs to nurture a 
product development culture that is similar to that of 
the commercial manufacturers whose partnership is 
sought Researchers must "speak the language” of 
industry to some extent. Several factors have been 
identified as useful for a successful product culture: 

• adopt a process model which takes into account 
marketing as well as technical development, and 
which reviews progress at identifiable stages; 

• identify product champions early in the projects; 

• draw upon the skiUs and experience of others who 
are closely allied to the organisation; 

• provide infrastructural support beyond the traditional 
clinical or research models. 



ERJC 220 



RESNA ’94 



^234 



June 17.22, 1994 



Ontario Rehabilitation Technology Consortium 



Stage 


Technical Activity 


Market Activity 


Product Documents 


Decisions 


Generate Idea 


Identify Ideas 


Identify Ideas 


Proposals, Sketches 


Screening 


Assess 


Technical 

Assessment 


Preliminary Market 
Assessment 


Simple Schematics, 
Nomenclature 


Preliminary 

Evaluation 


Define 

Concept 


Concept Design 


Identify, Test Concept 


Renderings, 

Instorctions 


Concept Evaluation 


Develop 


Product Development 


Marketing plan 


User Manual 


Evaluation 


Test 


Prototype Testing 


Prototype Field Test 


Dealer Support 


Evaluation 


Trial 


Final Design & 
Assembly 


Test Market 


Complete Product 
Description 


Financial Analysis 


Launch 


Full Production 


Market Launch 


User Support 


Post-launch 



Activities at Stages of New Product Development 



A product development process can be realised in 
seven stages (2). 1, Generate Ideas; 2, Assess; 3, 
Define Concept; 4, Develop Product; 5, Test; 6, Field 
Trial; 7, Market Launch. Each stage includes technical 
and marketing activity. A go/no-go decision separates 
each stage. Starting with broad freedom to generate 
many new product ideas, each succeeding stage is 
more cosUy and must therefore be more restrictive. 
The organisation should be prepared to shelve a 
product idea rather than develop it beyond what is 
reasonable from a cost/benefit standpoint. The 
documentation requirements of each stage are 
significant, as they form the basis of an effective 
technology transfer mechanism. This can make it 
easier for commercial collaborators to make their own 
assessments as to the value of a concept, and expedite 
the transfer of the technology to manufacture. 

In the table above, each stage of development is 
shown vertically. Columns indicate the kinds of 
workassociated with each stage, in categories of: 
Technical and Market Activity; Product Documents; 
and Decisions. 

The activity of identifying product or process 
champions is integral to the ORTC team structure. 
Strong commitment to the product concept can 
develop inside the Team Advisory Panels, where 



consumers, clinicians, marketers and researchers 
decide upon priorities, review projects and endorse 
new proposals. Consumer members of the Advisory 
Panels can help to draw other consumers in for focus 
groups, as well as bringing their own informed 
opinion into the design decisions as each project 
progresses. Similarly, members with marketing 
expertise can be influential from an early stage. 
Through their participation on the Management 
Committee, which reviews and approves projects with 
the guidance of the ORTC Advisory Board, team 
leaders have the opportunity to present on their own 
team’s activities, and to engage in constructive 
criticism of the other teams. 

The Consortium networking environment creates new 
opportunities for sharing skills and experience in 
technology transfer. First, each R&D centre within the 
Consortium has its own existing mechanisms for 
technology transfer, which can be enhanced and 
shared by the other Consortium members. Second, the 
tradition^ boundaries between types of disabilities are 
being lowered: the resulting cross-fertilisation of ideas 
and methods has proven to be a significant factor in 
several projects. Involvement in product development 
by Consortium members can follow several different 
routes: for example, one can work with fabricators to 
make a low volume, service intensive product to be 



, > 235 

RESNA’94 • June 17-22, 1994 



221 



Ontario Rehabilitation Technology Consortium 



used by one or several clinics; or work with industrial 
partners on a potential high return product, from the 
early idea assessment stages; or hand over 
development to an industrial partner at the end of 
concept definition. The choice of path depends on 
many factors, and it is important in managing 
technology transfer to focus on how each of those 
involved can benefit from following a particular route. 
In this regard, hindsight is 20:20. The opinion of 
someone who has made these decisions before, and 
who is committed to the organisation’s goals, can be 
invaluable. 

Studies of product development have identified the 
need for an organised plan, identifying discrete stages 
of the process, to contain costs and to demonstrate 
achievement of specific goals. A Technology Transfer 
Unit ( IT U), as part of the Consortium management 
support, can facilitate the attainment of product 
develc^ment goals. It can provide access to a shared 
base of experience and of resources for new product 
development, across the sfructure of scientific, 
engineering and psycho-social research in assistive 
device categories. Funds can be allocated on a cost- 
recovery basis to do work that has been identified as 
crucial to the commercial success of a well developed 
product concept. The potential for royalty paybacks to 
support further activities is a significant fector in the 
review process. The ORTC was fortunate in securing 
financial and advisory support at an early stage of this 
endeavour, from the National Research Council of 
Canada (NRCC). This federal government department 
continues to provide assistance for commercialisation 
through its Industrial Research Assistance Programme. 

RESULTS 

During its first two years a provincially funded 
programme of rehabilitation research and development 
has evolved a strong focus on new assistive devices. 
Several products which were on the market at an 
earlier date have been upgraded significantly; 
introduction of new products which otherwise would 
not have been developed, is now imminent; and 
licensing agreements for new assistive device 
concepts have resulted from the concerted efforts of 
several of the ORTC teams, and of the Technology 
Transfer Unit 



REFERENCES 

1. Milner, M. "The Ontario-based Rehabilitation 
Technology Research and Development Consortium." 
RESNA ’92, Proceedings of the 15th Annual 
Conference, p.405. 

2. Cooper, R.G. Winning at New Products. Holt, 
Rinehart and Winston, 1986. 

3. McNeal, DJI. "Commercialization Activities of 
Rehabilitation R&D Centers", a NIDRR study, 
available upon {^plication to Rancho Rehabilitation 
Engineering Program, 7503 Bonita St., Downey, CA, 
telephone: (310)940-7994. 

4. Knorr, V.C. "Commercialization, a Natural 
Extension of Technology Transfer". ICAART 88, 
Proceedings of the 11th Annual RESNA Conference 



ACKNOWLEDGEMENTS 

Base funding for the ORTC is provided by the 
Ontario Ministry of Health. The Technology Transfer 
Unit was initiated with a contribution by the National 
Research Council of Canada. 



William (Bill) Bennett 
Rehabilitation Engineering Department, 

The Hugh MacMillan Rehabilitation Centre, 
350 Rumsey Road, 

Toronto, Ont CANADA M4G 1R8 
(416)425-6220, fax:(416)425-1634 
E-mail: rtrdcbb@oisel.oise.on.ca 





RESNA ’94 • June 17-22, 1994 



222 



ASSISTIVE TECHNOLOGY TRAINING: AN INTERDISCIPLINARY APPROACH 



Richard Baer, Marvin Fifield, Amy Henningsen, Linda Chisholm, Beth Foley 
Center for Persons With Disabilities - Utah Slate University 
Logan, Utah 84322-6800 



Until recently, there has been minimal 
cooperation between technical and 
human service disciplines in the 
development and provision of assistive 
technology (AT) and virtually no 
interdisciplinary training programs 
(TIP). In 1992 a ITP in AT was 
initiated at Utah State Univ^ity. 

The present pap^ describes the program, 
including disciplines represented, 
training experiences, AT products 
developed by students, and student 
evaluation of the program. 



Background 

The science and practice of developing 
devices and slices to assist individuals 
with disabilities is refmed to as assistive 
technology (AT). Traditionally, human 
service personnel such as occupation 
therapists, vocational rehabilitation 
counselors, special educators, and speech 
pathologists sought to obtain AT devices for 
the consumes with whom they worked. 
Many of these devices were developed for 
other purposes by persotmel firom technical 
disciplines such as mechanical engineering 
or computer science.. Later they were 
adopted or adapted for use by individuals 
with disabilities. Until recently, formal 



interdiscqrlinary cooperation betwera human 
service and technical disciplines in the 
development of AT has been extremely 
limited, and interdisciplinary training 
programs have been virtually nonexistent 

Objective 

Develop an intndisciplinary training 
program in AT that marries technical and 
human service expertise in the development 
and delivoy of AT, and increases the pool 
of available AT posonnel. 

Approach 

Recognizing the need for increased 
cor^ration among human sovice and 
technical discqrlines in the development of 
AT, the Center for P^ons With Disabilities 
(CPD) at Utah State Univ^ty initiated an 
interdisciplinary training program in 1992. 
Seniors tom a variety of disciplines, 
including electrical and mechanical 
engineering, industrial technology, computer 
science, corrununicative disorders, social 
work and special education, were recruited. 

Students in the program participate in a 
variety of learning experiences. Through a 
series of weekly seminars they learn basic 
information about disabilities and assistive 
technology. Guest speakers include 
designers, manufacturers, and suppliers who 
give students a real world view of AT 
development and provision. Meetings with 
consumers allow students to obtain first 
hand information about disabilities and a 
personal perspective on individual needs for 
assistive technology. Through field trips 
and laboratory work students have opportun- 
ities to become familiar with a wide variety 
of devices. They also participate on inter- 
disciplinary teams that work directly with 
hospitals, schools, nursing homes and 
independent living centers. These teams 




RESNA ’94 • June 17-22, 1994 



223 



Inter. Trgn. in AT 



ad£q)t existing technology, design new 
technology, train people to maximize their 
use of technology and identify funding 
sources for technology. 

Results 

Detailed syllabi outlining specific training 
experiences woe developed for the three 
quarter sequ^tial duration of the inter- 
disciplinary training program. Students 
rated their satisfaction with and the value of 
each expmence, and commuted on how 
they might be improved. In gen«al, the 
ratings were very high. Detailed data will 
be provided and discussed as part of the 
interactive paper session. 

Intodisciplinary student teams developed 
numerous innovative AT products for the 
consumes they worked with, e.g. a 
reasonably priced communication device 
incorporating new ccanputer technology, a 
battoy-powraed toy car adtqtted with a joy 
stick that made a two year old with cerebral 
palsy indepradently mobile, etc. 

Photographs and detailed descr^tions of 
these devices will be provided as part of the 
interactive psgter session. 

Discussion 

Students are involved in a practical into'- 
disciplinary team tqjproach which enables 
then to learn from each othe'. At the same 
time, consumers beiefit fiom the students’ 
combined efforts. Students fiom technical 
disciplines develop increased awareiess of 
disabilities and broader their information 
base for making career and community 
service choices. Students from human 
service disciplines gain first hand expeience 
with technology that will benefit the 
consumers they woric with throughout their 
carees. 



Acknowledgements 

U.S. Department of Education - National 
Institute on Disability and Rehabilitation 
Research 

Department of Health and Human Sovices - 
Administration oa Develoimiratal 
Disabilities 

Dr. Richard Baa* 

Center for Persons Witii Disabilities 
Utah State Univosity 
Logan, UT. 84322-6800 
(801)750-7009 
(801)750-3944 FAX 





RESNA’94 • Jane 17-22, 1994 



224 



RERC ON TECHNOLOGY EVALUATION AND TRANSFER: 
PROGRAM ACCESS AND VALUE ADDED 

Joseph P. Lane, MBPA 

Rehabilitation Engineering Research Center on Technology Evaluation and Transfer 
Center for Assistive Technology, University at Buffalo 



ABSTRACT 

The RERC on Technology Evaluation and 
Transfer is a collaborative program involving 
consumers coordinating user trials, businesses 
performing market analysis, and a university- 
center conducting technical evaluations. The 
RERC is designed to add value to potential 
products for the assistive technology 
marketplace. Inventors, researchers and 
companies have multiple access points to the 
RERCs capabilities. 

STATEMENT OF THE PROBLEM 
A collaborative solution to the major problems in 
the assistive technology marketplace was 
reported previously (1). This collaborative 
solution is now implemented through the 
Rehabilitation Engineering Research Center on 
Technology Evaluation and Transfer (RERC- 
TET), at the University at Buffalo. 

The RERC-TET follows the principles of 
Participatory Action Research by involving 
stakeholder in all aspects of the process (2). 
Researchers who develop and test, business 
people who market and sell, and consumers who 
purchase and use assistive technology, all 
participate in the evaluation and transfer of new 
assistive devices. 

Also, the RERC-TETs long-range plan takes the 
policy of inclusion to a logical conclusion. The 
program will eventually become a community- 
based enterprise, directed by and for persons with 
disabilities, with continuing technical support 
from the research center, and mariceting support 
from the business community. A not-for-profit 
corporation named AZTECH is the basis for this 
conununity venture. 

This paper describes the RERC-TETs program, 
presents points of access for perspective users, 
and explains how the RERC-TETs capabilities 
add value to new assistive devices. 



APPROACH 

The RERC-TETs mission is to help prototype 
assistive devices reach the marketplace by 
demonstrating their value through technical, 
consumer and market evaluations, and by 
identifying commercialization partners willing to 
turn the prototype into a product for the 
marketplace. Anyone with a prototype assistive 
device is eligible to participate. Inventors, 
researchers and companies are invited, while 
persons with disabilities are especially 
encouraged to participate. 

The RERC-TET has a five phase model for 
performing this mission. For a small application 
fee, the project team will assess a prototype's 
commercial potential by examining its technical 
function, consumer value and market potential 
(Phases I and II). If the prototype shows 
promise, the RERC-TET will negotiate an 
agreement with the inventor to conduct further 
evaluations and seek commercialization partners 
(Phases III, IV and V). These additional phases 
are conducted at the sole discretion of the RERC- 
TETs management team. At the same time, the 
inventor can decline to participate further. These 
five phases are described below. 

Phase L Intake and Screen. Phase I is the 
point of contact for The RERC-TET. Phase I 
determines if the contact from an inventor, 
researcher or corporation fits the program. For 
example, the RERC-TET only works with new 
ideas developed into a prototype device. Callers 
with only ideas for products are referred to 
national programs working in those specialty 
area of technology. Callers seeking investment 
in a commercial product, or callers with 
inventions that clearly duplicate devices already 
in the marketplace, fall outside the project's 
scope. Callers in Phase I who appear to have a 
new prototype assistive device, are invited to 
complete a questionnaire on the device's 
background and submit it along with the 
prototype device as a candidate for the Phase 11 
evaluation. 



^ 239 

RESNA’94 • June 17-22, 1994 



225 



RERC ON EVALUATION AND TRANSFER 



Phase II. Concept Review. Phase II involves a 
fairly intensive review of the prototype device, 
focusing on the device's potential value to 
consumers. Potential value is emphasized over 
present value because the prototype submitted 
may lack sufficient refinement to immediately 
demonstrate its full value as a future product. 

Phase II has two steps. The first step is an initial 
review by a standing committee of a 
rehabilitation engineer, a business analyst and a 
consumer. This committee examines the 
prototype device and supporting documentation, 
to determine if the device is technically feasible, 
has functional utility, and is different from 
current products. If not, the application, 
prototype and fee are returned to the inventor 
with a referral to other resources. 

In Phase IPs second step, the marketing team 
investigate the device's ownership, development 
history, patentability and marketability; the 
technical team evaluates the devices fimction, 
safety and operating requirements; and the 
consumer team conducts focus group interviews 
on functionality, usability and appropriateness. 
The RERC-TET also searches for a 
commercialization partner, without revealing any 
proprietary information about the device. 
Responses provided important feedback about 
the device's value in the marketplace. 

Phase II concludes with the inventor receiving a 
summary report from these analyses, including a 
recommendation about the device. This report 
concludes the RERC-TETs obligation for the 
application fee. Additional work is performed 
under a separate agreement with the inventor. 

Phase in. Research and Development. If the 
prototype device demonstrates conceptual 
potential but still needs significant modifications 
to meet the needs of marketplace, it is considered 
an alpha prototype. The RERC-TET may 
propose an agreement to perform the work 
necessary to generate an improved version of the 
device considered a beta prototype. Developing 
a beta prototype typically involves initiating any 
appropriate legal protection over the device, 
performing the marketing and technical work 
outlined in the Phase II evaluation, verifying the 
device's safety, then involving multiple 
consumers in pilot testing of short duration. 




The research and development work is a 
substantial investment of time and resources. 

The RERC-TET may stop the Phase III work at 
any point where the device demonstrates an 
insurmountable deficiency. Devices developed 
into beta prototypes that receive positive reviews 
by consumers, are then presented to potential 
commercialization partners. Either an external 
organization will agree to commercialize the 
device, or their feedback may prompt additional 
modifications. 

Phase III concludes with the beta prototype and 
supporting documentation available for 
commercialization. The RERC-TET and the 
inventor will then decide to continue working 
together or end their partnership. 

Phase IV. Prototype Evaluation. Devices 
developed to the beta prototype level are eligible 
for Phase IV. These devices may come directly 
from Phase II, or first move through Phase III. 
Prototype evaluation requires first constructing 
multiple copies of the device, then conducting 
extensive evaluations to determine how well the 
device functions as a consumer product. 

For example, the technical team uses multiple 
copies to tear down, bum in and wear out. The 
consumer team uses multiple copies to conduct 
extended (multiple week) user trials involving 
test sites representing various demographic, 
geographic and climatic conditions. The 
marketing team uses multiple copies in 
presentations to potential investment or 
commercialization partners. 

Phase IV should end with a transfer of the 
assistive technology to an outside entity. If 
transfer does not occur but the RERC-TET is 
convinced the device has adequate value for 
consumers, the final option is Phase V. 

Phase V. Device Production. The RERC- 
TETs community-based business entity, 
AZTECH, may elect to product a device locally. 
In this case, AZTECH would pursue the 
resources to capitalize and implement a start-up 
company. Phase V is reserved for devices 
serving an important function but that cannot 
reach Ae marketplace any other way. 



226 



RESNA ’94 



June 17-22, 1994 



RERC ON EVALUATION AND TRANSFER 



IMPLICATIONS AND DISCUSSION 

The RERC-TET identifies, secures and 
communicates information on the functional 
value of new devices. Most inventors lack access 
to the technical, business and consumer networks 
capable of delivering this information. 

Access to the RERC-TET. Any individual 
inventor or researcher, and any corporation has 
direct access to this program via telephone, fax 
or mail. If the program does seem appropriate, 
the inventor is sent an application package which 
is available in alternative access formats. The 
inventor is asked to return the application 
package and a functioning prototype with 
minimal delay. The inventor must also submit an 
application fee of fifty dollars, although a waiver 
is available for applicants with limited incomes. 

If the RERC-TET determines that the submission 
is not appropriate for whatever reason, all 
materials and the fee are returned to the inventor. 
The program also provides referrals to other 
agencies that may be able to help the inventor 
develop their concept or device. 

The RERC-TET conducts the entire Phase I and 
Phase II process at no additional cost or 
obligation to the inventor. The Phase II 
information helps everyone realistically appraise 
the potential value of the prototype invention. 

The RERC-TET team and inventor jointly 
determine if the device moves to the later phases. 

Researchers, corporations and consumer agencies 
can also access the RERC-TET by participating 
in the evaluation programs. Every additional 
participant strengthens the national information 
network and presents additional opportunities for 
collaboration. The RERC-TET expects to rely 
heavily upon the combined expertise of the 
research, development, training and information 
programs funded through the National Institute 
on Disability and Rehabilitation Research. 

Value Added by Program. Contrary to popular 
myth, building a better mousetrap is no longer 
sufficient to draw the world's attention. Forty 
years ago, the marketplace had adequate capital 
to underwrite the cost.of implementing good 
ideas. In the recent past, the marketplace was 
still willing to gamble on promoting a new 
product once the idea was reduced to practice. In 
the present economy, proof of concept is A 



replaced by proof of market criteria. The burden 
is now on the inventor to demonstrate that a 
market exists for a new device. Any information 
that helps demonstrate that market enhances the 
potential value of the new device. 

Every phase in the RERC-TET's program is 
designed to add value to a prototype assistive 
device. For example, the initial screening 
determines if an invention qualifies as an 
assistive device, and identifies competing devices 
already in the marketplace. The Phase II 
Concept Review generates information about the 
device's technical and functional capabilities and 
target markets, and elicits direct consumer and 
marketplace reactions to the device. 

The Phase III Research and Development activity 
transforms a rough prototype into a pre- 
production model, by incorporating the 
modifications identified in Phase II and testing 
those modifications under controlled local 
conditions. The Phase IV Prototype Evaluation 
work verifies prior work through nation-wide 
testing under all anticipated conditions. Phase V 
actually initiates production, which hopefully 
generates sufficient sales to generate interest 
from other firms in the marketplace. 

The RERC-TET is available to pursue any 
arrangement that makes useful assistive devices 
more available to persons with disabilities. 

REFERENCES 

1. Lane, J. (1993). "A Collaborative Model for 
Technology Evaluation and Transfer." In J.J. 
Presperin (Ed.) Proceedings of the RESNA '93 
Annual Conference. Washington, DC: RESNA 
Press. 225-227. 

2. Whyte, W. F. (Editor), \99\, Participatory 
Action Research, Newbury Park: Sage 
Publications. 

ACKNOWLEDGMENT 

This work is supported by The National Institute 
on Disability and Rehabilitation Research, U.S. 
Department of Education. 

Joseph P. Lane, MBPA 

Center for Assistive Technology 

University at Buffalo, Buffalo, NY 14214 



RESNA ’94 • June 17-22, 1994 



227 



THE VA REHAB R&D TECHNOLOGY TRANSFER PROCESS 



Saleem J. Sheredos, Mary E. Cupo, and James M. Ford 
Department of Veterans Afiairs, Rehabilitation Research & Development Service, Technology Transfer Section 

Baltimore, Maryland 



O 

ERIC 



ABSTRACT 

The Department of Veterans Affairs (VA), 
Rehabilitation Research & Development 
Service’s (Rehab R&D) Technology Transfer 
Section (TTS) screens the products emerging 
from rehabilitation R&D, primarily sponsored 
by the V A. Requests involvingnon-VA funded 
development are also reviewed to identify 
products or techniques that may meet specific 
VA needs within one of the priority areas: 
Prosthetics/Amputations/ Orthotics; Spinal 
Cord Injury; Communication, Sensory, and 
Cognitive Aids; and. Aging. The TTS is 
responsible for the design and management of 
a systematic process to validate proven 
rehabilitation R&D findings and progressing 
the successful outcomes into clinical use, 
product manufacture, and commercial 
availability. The ultimate goal is for timely 
transition of prototypic development into 
commercially viable products and techniques 
that can be readily available to benefit veterzins 
and non-veterans with disabilities. Through 
national information dissemination of 
evaluation results, the TTS contributes 
knowledge about new products and techniques. 

BACKGROUND 

Technology transfer has been defined many 
ways. Much of what is understood about the 
process involves the exchange of information 
(1). Most technology transfer programs tend 
to focus on technology licensing and 
cooperative R&D which is understandable since 
these two activities involve transfer of 
intellectual properties (2). This paper describes 
a method of technology transfer that facilitates 
the progression of products borne out of the 
R&D arena into an objective environment that 
incorporates aspects of manufacture, 
evaluation, and, if deemed appropriate, 
availability in the marketplace. Based upon 
this process, technology transfer is viewed as a 
transformation of knowledge into successful 
new products and techniques (3) that are 
affirmed to be functional, safe, and effective 
for use by veterans and non-veterans with 
disabilities. 

OBJECTIVE 

The mission of the VA Rehab R&D is to support 
an intramural R&D program for improving the 
quality of life of veterans with disabilities. Non- 
VA fimded requests are also reviewed according 



to selection criteriaforacceptanceasatechnology 
transfer project. The end result is realized in the 
av 2 ulability of products and techniques that can 
provide greater functional independence for the 
veteran andnon-veteran populations. Tosiq>port 
this mission and provide for prompt transfer of 
promising products and techniques into 
commercial production and clinical use, the 
Rehab R&D established the TTS. 

To fulfill its primary role of making effective 
products and techniques available to veterans, 
andothers,whomaybenefitfiomthetechnology, 
the TTS employs a systematic process that 
involves the developer, a manufecturer, VA 
Central Office (VACO) Service Director(s), and 
clinical test sites. 

METHOD 

Activities and phases of the technology transfer 
process will be discussed using the schematic 
illustration on the next page. 

Once the research idea/concept (R) has moved 
into development (Dl), the outcome is usually a 
working prototype. Internal testing and 
evaluation to demonstrate the feasibility of the 
new (alpha) device or technique is accomplished 
by the R&D principle investigator (PI). Design 
changes to the alpha model may be necessary 
based upon incoming data fiom ^e developer’ s 
testing. The working prototype (D2), or alpha 2 
model, would have completed successful 
laboratory and limited clinical trials prior to 
entering the technology transfer process (D3). 

It is at this juncture that the R&D PI submits a 
Request For Evaluation (RFE) to the TTS. The 
RFE elicits specific information that is used to 
review the appropriateness and readiness of the 
development as a TTS project. Receipt of a 
completed RFE officially commences the 
technologytransferprocess. This is followed by 
a peer review using the following selection 
criteria (Gateway): 

• VA level of need/level of interest: 

di) Product/technique classifies within one of 

Rehab R&D priority areas - 

1) Prosthetics/Amputation/Orthotics; 

2) Spinal Cord Injuiy (and other disabling 
neuromuscular disorders); 

3) Communication, Sensory, and Cognitive 
Aids; and 

4) Aging. 



RESNA ’94 • June 17-22, 1994 



228 



TECHNOLOGY TRANSFER PROCESS 



'Veterans/ VA Clinics ' 



NeeJLMd, — I 



Technology Transfer 



Initiate Tech Transfer 

® 



Seteotion Criteria 

VA Need 
Fit for Use 
Manufacturable 
Marketable 



mm 



Research \ 

{r) I deas (ft Start 







:Mfg Beta Modeto 
:Hlot Cvei 
n or 2 Sites) 






Design Changes 
From Beta Test 
■.Freeze Design 
:Modity or Construct 
New Beta Models 
if Reqiired 




Comnterciaff2ation 

Multiple Product Changes as a Ftesult 
of Consuner FeedbacK Throughout 
the Life Cyde of the Product 



Field Evaluation: Design Changes 
(3-50) Freeze Design 
1st Commercial 
Production Run 
Marketing Strategy 
Training & Education of Mfg Staff 
B.g., marketing, field si4>port 



V® Alpha Test Completed 
Successful Lab & Umited 
Cbnstmer Use Data 



Development 



© End InvolvoTtent of RfiiD 



Concepts 
Design Goals 
Ccmputer Simulation 
Basic Sdentific Information 



Development 



t 



Transition Point 
Occurring at Anytime 
Between Phases 



C = Commerdai Phase 
D = Deveiopment Phase 
R = Research Phase 
T = Technoiogy Transfer Phase 
♦ = Decision Point - Go/No-Go 



b) Product /techn ique should improve 
treatmenUnanagement and rehabHitatioD 
of veterans as identified an d verified bygone 
or more of the following: 

1) Appropriate VACO Service Directors; 

2) VA Service Chiefs and/or clinicians 
recommended by Service Directors; 

3) Veteran Service Organizations; and 

4) Available VA statistics. 

• Fit For Use 

Alpha testing successfully completed on 
Majdag- prPtQtypg(^) to inclpd^.;. 

1) Lab testing against recognized 
standardized tests for the following: safety 
issues and performance specifications. 

2) Successful (limited) consumer use: 
Minimum clinical data (1 -2 subjects) 
from developer/ manufacturer ( If all 
other criteria are met, this category usually 
requires a Pilot Evaluation); Adequate 
clinical data (3 or more subjects) from 
developer/manufacturer (If all other 
criteria is met, this category usually leads 
to a national, multi-center evaluation. 

• Manufacturable/Pre-Commerdal 

a) Ready for manufacturing pre-commercial 
fheta'i models as demonstrated bv: 

1) Adequacy ofperformance specifications, 
function^ description, and illustrations. 

2) Legal issues resolved or in progress i.e., 
patent rights, license to manufecture, 
CRADA, publicdisclosure, and royalties. 



RESNA ’94 



• Marketable 

a) Market potential is judged in terms of the 
following : Percentage of target population 
(veteran vs non- veteran) that is likely to benefit 
from actual use of the product or technique; 
Competitive products vs major advantages of 
developed product; Cost to manufacture and 
market price; Training requirements to 
successfully transfer product or technique to 
routine use; and Service requirements. 

Once the RFE peer review is complete and 
responses are positive, the TTS will formulate and 
submit a recommended plan of action, including 
budget support, to the Director, Rehab R&D 
Service. Approval at this level commences the 
manufacture and evaluation phases (Tl). A 
manufacturer is identified for the initial tooling, 
and engineering required for fabrication of the 
first production prototypes. As part of the 
procurement contract, the manufacturer is required 
to provide a statement of commitment to market 
the product or technique pending successful 
outcome of the evaluation. Allocated funds are 
transferred to the contracting VA Medical Center 
for manufecture of pre-commercial (beta) models. 

Upon completion of manufacture, the beta testing 
and evaluation (T2 -T4) can commence. Based 
upon the extent of clinical data reported in the RFE 
by the R&D PI, a pilot study (approx. 3 months 
duration at 1 or 2 sites) may be necessary in order 
to generate additional data to support proceeding 
to a national, multi-center evaluation. The TTS 
collaborates with the appropriate VACO Service 
Director for selection of participating evaluation 
sites within the VA. The objective of the clinical 

.V- -243 

• June 17-22, 1994 229 



available 



TECHNOLOGY TRANSFER PROCESS 



evaluation is to validate the following areas: safety, 
effectiveness, reliability, benefits for the intended 
user, and the degree of commercial readiness of the 
product or technique. Depending upon the 
complexity of the project, it is oftentimes required 
that TTS project manager make several on-site 

visits to the^participating medical center to review 
protocol requirements for implementation, and 
provide staff training. Ehiring the course of an 
evaluation, data may indicate a technical problem 
which will necessitate a major desi^ change in 
the beta model. It is at this juncture that a “freeze” 
is put on the evaluation. The developer and 
manufacturer must decide upon the appropriate 
modification(s) to resolve the problem. TTie revised 
beta models are then redistributed for continuation 
of the evaluation. The revision(s) would be 
incorporated into the conunercial version. For 
example, at mid-point of an evaluation of the 
Synergetic Prehensor (a myo-electrical controlled, 
powered upper extremity prosthesis) subject data 
indicated a problem with grip force and reliability. 
A recall of the beta models was performed to 
incorporate a redesigned backlock mechanism to 
improve pinch force and minimize jamming. 
Evaluation resumed and final subject data showed 
that the retrofit significantly reduced problems 
associated with jamming and pinch force. As a 
result of the information obtained in this study and 
the manufacturer’s redesign, a number of 
improvements have been made to the production 
model. 

The resultant evaluation data is analyzed by TTS 
and a recommendation is generated and submitted 
to the VACO Prosthetic Technology Evaluation 
Committee (PTEC) for final review and approval. 
If recommendation is to proceed to 
commercialization and receives PTEC approval, 
the TTS forwards this information and copy of 
final reporttotheOfficeofAcquisitionandMateriel 
Management, VACO, and itsNational Acquisition 
Center in Hines, Illinois. The manufacturer also 
receives notification and copy of the final evaluation 
report. 

It is anticipated, especially in the early conunercial 
years but also through the lifetime of the product, 
(Cl - C3) that consumer use will result in further 
refinement to continue to improve fimctional 
advantages and reliability. New technology that 
fills a void may be deployed to select VA facilities 
to encourage its integration into practice. The TTS 
also disseminates information concerning the 
commercial availability of successfully evaluated 
products and techniques through Product 
Evaluation Data Sheets, publication of articles in 
professional journals, and presentations at 
conferences. 



IMPLICATIONS AND DISCUSSION 

For effective technology application there needs 
to be a mechanism in place that not only supports 
the research and development but is also able to 
bring to fruition proven, successful R&D outcomes 
that are ready to be accessed by persons with 
physical and/or sensory impairments. It is realized 
that this is not a simple but complex process that 
requires a well-managed, purposeful plan of action. 
As part of the VA Rehab R&D Program, the TTS 
orchestrates an active approach that incorporates 
manufacture and evaluation (subject usage) early 
in the planning phase thus fulfilling two important 
areas: 1) “locking in” a manufacturer committed 
to supporting the product or technique in the 
commercial market and 2) gleaning clinical 
evaluation data from an objective environment 
that can be used to validate and enhance fimction 
and reliability of the product or technique. This 
process affirms user benefits and eq^plication, and 
generates amaxketopportunity for the final product 
or technique. Only through conceited efforts in 
networking and communication by those who 
effect varying technology transfer approaches 
and models can we remain focused on die exigent 
need for improving technology transfer practices 
and increasing the availability of commercial 
products designed to ameliorate the lives of persons 
with disabilities. 

REFERENCES 

1 . McCardle,K.F. (1983). InformationAcquisition 
and the Adoption of New Technology. 
Management Science, Vol. 31, No. 11, pp. 
1372-1389. 

2. Herrmann, J.F. (1983). Redefiningthe Federal 
Government's Role in Technology Transfer. 
ResearchManagement, Vol. 26, No. 1, pp. 21- 
24 

3. Camp, S.M. and Sexton, D.L. (1992). 
Technology Transfer and Value Creation: 
Extending the Theory Beyond Information 
Exchange. Journal of Technology Transfer, 
Vol. 17, Nos. 2&3 (Double Issue) 

Saleem J. Sheredos, Rehabilitation Engineer 
Program Manager 

VA Rehab R&D Service 
Technology Transfer Section 
103 South Gay Street 
Baltimore, Maryland 21202 ^ 

(410)962-2133 



ERJC 



230 






RESNA ’94 



17-22, 1994 



•• •* 







PRODUCT DESIGN FOR PEOPLE WITH DISABILITIES: 

A NEW GRADUATE COURSE AT THE UNIVERSITY OF TENNESSEE MEMPHIS 

Michael J. Rosen 

Biomedical Engineering Department 
University of Tennessee Memphis 



O 

ERIC 



ABSTRACT 

A new graduate course on Product Design for People 
with Disabilities has been developed in the 
Biomedical Engineering Department at the University 
of Tennessee Memphis campus. It is a project course 
which requires the design and prototyping of a device 
or system to meet a specified need related to a 
particular disability and activity. While focused on 
assistive technology, it is meant to teach lessons 
regarding design of human -interactive products which 
are generalizable to mass-market devices. Its lecture 
content covers a broad range of topics related to 
disability, design methods, and relevant technologies. 
The location of the BME Dept, on a "medical 
campus" offers the advantage that end users of 
assistive technology and expert providers of 
rehabilitation services are on hand to provide lectures, 
consultation, and design evaluation. 

INTRODUCTION TO THE COURSE 

Product Design for People with Disabilities (PDPD) 
is characterized by an unusual combination of 
features: 

• In a single academic semester, students design and 
prototype working assistive technology to solve a 
problem which has not been solved by current 
commercial products. 

• Students work in teams and are graded according to 
team performance rather than individual contributions. 

• The course is taught in classroom and shop space at 
the UT Rehabilitation Engineering Program. This 
provides the course students with complete 
immersion in an environment populated by potential 
end users of their work and service providers — as 
well as the engineering faculty. 

• Outside experts including users, practitioners, payers 
and potential manufacturers are directly involved in 
the evaluation and refinement of students' concepts 
and designs throughout the course. 

• Lectures from several BME faculty members and 
visiting speakers provide the disciplinary breads 
necessary to teach topics as broad as rehabilitation 
engineering and design. 

• Design projects are chosen in part for their 
integrative pedagogical value; they require students to 
combine what they've learned in class work — with 
little regard for disciplinary boundaries — as well as 
their practical experience as users of technology and 
manipulators of the physical world. 

• Marketing, legal and entrepreneurial issues are also 
important parts of the syllabus. 

b 



BACKGROUND 

PDPD is being offered for several reasons. Most 
obviously, the Biomedical Engineering Department at 
UT Memphis is in the process of developing a 
graduate curricular track and research focus in 
Biomechanics and Rehabilitation Engineering (BRE). 
Three members of the faculty, including one with an 
endowed chair in rehabilitation engineering, focus 
their teaching and research in this area. To 
accommodate the interests of a variety of students, 
current and future course offerings will cover the 
range from reductionist description of human organ 
systems and their modes of failure to design of 
marketable technology to support function in 
activities of daily living. PDPD addresses the latter 
part of the spectrum of BRE activities. 

In addition, the BME Dept, as a whole is placing 
unusual emphasis for an academic program on 
conducting and teaching biomedical engineering 
design. For example, in addition to BRE, several 
BME faculty in the area of medical imaging are 
developing i^amatically improved approaches to 
making x-ray images. In engineering graduate 
programs, including biomedical engineering, an 
emphasis on the practice of design as an academically 
worthy endeavor continues to be uncommon. PDPD 
is one of the current or planned courses which 
establish this department's commitment to teaching 
biomedical engineering design. 

Further, this course provides one of the major 
mechanisms for interaction between the UT Memphis 
Rehabilitation Engineering Program (UTREP), of 
which the author is the Director, and BME students 
and faculty. The therapists and technical staff of the 
Program deliver assistive technology services in the 
areas of seating, computer interfacing, augmentative 
communication and job accommodation. They and 
the consumers they serve in effect provide an in-house 
informal consultation service for the PDPD students, 
one which is typically impossible to obtain on 
engineering campuses. Their involvement in 
problem formulation and design evaluation insures 
that the students' designs are constrained and informed 
by the "voice of the customer". Accessibility of this 
community to the students is enhanced by conducting 
both the classroom work and fabrication activities in 
the UTREP building. 

From the perspective of the Rehabilitation 
Engineering staff, the opportunity to teach and advise 
engineering graduate students also has value. These 
students' knowledge of technology and the perspective 
from which they ask questions can lead to the 



,245 



RESNA ’94 17-22, 1994 



231 



Product Design Course at UT Memphis 



formulation of new ideas for assistive technology and 
rehabilitation service delivery. This has value to the 
Program since conducting r&d directed at advancing 
particular areas of assistive technology is a major part 
its mandate. Some of the same students who become 
involved with UTREP through the course will be 
conducting their thesis projects in rehabilitation 
engineering via grants awarded to the Program. 

The greater Memphis industrial community also 
offers a supportive environment in which to teach 
PDPD. This area is home to Smith and Nephew 
Richards, Danek, Wright and Shering Plough, all of 
which develop and manufacture products for the 
practice of orthopaedic rehabilitation. Existing 
contacts between members of these firms' research 
staffs and members of the BME faculty facilitate 
informal involvement of these companies in the new 
course. In particular, advice and lectures from their 
staff on the constraints imposed on design by 
manufacturing, marketing and medical device approval 
procedures is readily available. 

The conviction that small groups of talented but 
inexperienced students can make tangible progress in 
solving challenging rehabilitation design problems in 
a short period of time is based on the experience of 
the BME faculty with related teaching efforts at other 
universities. In particular, the author regularly 
introduced assistive technology projects into the 
required Mechanical Engineering Department senior 
design course which he taught at MIT. In one term, 
for example, each of four teams of twenty five 
students designed and prototyped completely new 
equipment for equestrians with paralysi. One example 
was a mount-dismount lift, specialized saddle, and 
joystick interface to the reins for quadriplegic riders. 

DETAILS OF COURSE CONTENT AND 
FORMAT 

In its present form, salient details of the course 
include the following: 

• There are two lectures and one recitation section 
each week during a fourteen-week term. 

• The project is defined only in terms of a desirable 
activity and a specified disability. Defining what the 
problem is and narrowing the scope of acceptable 
solutions, i.e. the phase which comes before 
conceptual design, is an essential part of the process 
the students are expected to experience. 

• There are three major deliverables: a list of the 
design goals in the language of the intended users and 
a matrix which translates these into engineering 
variables; a proposed selection of team's best 
conceptual design represented as drawings and text; 
and a first-pass operating prototype of the device 
presented and documented for a feculty and 
community audience. 

• Students are organized into independent teams of 
between 8 and 16 students, depending on course 
enrollment. Each team undertakes the complete 



assignment and all its members receive the same 
grade. 

• Upon submission of the conceptual design from 
each team, a rapid evaluation is conducted by outside 
experts including potential users. Their comments 
become the basis of the next lecture and are used to 
guide design revisions prior to detail design and 
prototyping. 

• Lecture content typically includes design methods, 
e.g. brainstorming, concept comparison charts and 
"QFD" methods for translating customer preferences 
into engineering variables; estimation for feasibility 
determination; graphical communication; a survey of 
current assistive technology and disability categories; 
generic design topics including reliability, safety, 
liability and patents; modeling of relevant areas of 
human function to guide design decisions; 
prototyping methods; group management methods 
and tools; and technologies relevant to the specific 
project. 

• At this writing, the project in the Spring of '94 is 
the design of a "room-cruiser" for non-ambulatory 
toddlers with neuro-muscular disabilities. It is 
intended to utilize any movement the child can 
produce to provide propulsion on a flat floor. It 
facilitates exploration of his/her environment despite 
the absence of coordinated lower limb motion. 

RATIONALE 

The central notion of desip project courses is 
not new and the rationale which applies in traditional 
disciplines appears to apply as well in rehabilitation 
engineering; design is a complex iterative process 
made up of numerous skilled creative and analytical 
activities. The process can be learned and 
personalized only through experience. This 
observation also provides one rationale for carrying 
the experiential learning of design through the 
prototyping phase. Another is the powerful 
motivational value, invariably noted by students, of a 
tangible outcome which can be demonstrated — not 
just drawn or described — to others. 

The decision to organize the course around 
student teams was based on the observation that that's 
the way product development actually happens in 
industry. A typical undergraduate engineering 
education places enormous emphasis on individual 
accomplishment and encourages competition among 
individuals. This offers poor preparation for working 
in industrial teams, especially given the current 
emphasis on "concurrent design" and the 
organizational demands of shared responsibility for 
quality. 

In presentation of the senior design course at 
MIT, it was found that the importance of sound team 
organization for the success of a product is also 
emphasized by requiring the fabrication of a 
prototype. Failure of team decision-making and 
coordination results in design flaws which are much 



RESNAr’94 24fiie 17-22, 1994 



232 



Product Design Course at UT Memphis 



more apparent when the design is brought to the 
point of physical implementation. 

The particular project chosen for the current 
presentation of the course was meant to exploit 
UTREP expertise in seating and mobility. The 
"reality** of the project for the students is also 
enhanced by the possibility that it could result in new 
technology — perhaps further developed with grant 
funds — which would be prescribed experimentally 
and evaluated as part of Program service delivery. 

The BME Department must meet the educational 
needs of those of its students who are motivated 
primarily by an interest in design of medical and other 
human-interactive products rather than a narrower 
focus on disability-related applications. Further, 
many of those students who do currently specify 
rehabilitation engineering as their intended 
professional specialty are likely to change tracks more 
than once in their careers and would therefore benefit 
by a course with broader applicability. For both of 
these reasons, it is important to note that reflection 
on the author’s experience using assistive technology 
projects in mechanical engineering design courses 
suggests that there are several ways in which design 
of assistive technology effectively conveys many 
generalizable lessons. These have been discussed at 
greater length elsewhere [1] but a few are excerpted 
here: 

• A real design task is almost never like "the problem 
at the end of the chapter"; it demands application of a 
broad range of disciplines and arts because that’s they 
way the natural world and the marketplace are. This 
multidisciplinarity is particularly evident in design of 
assistive technology. Design of adapted control 
interfaces for computers, for example, had better take 
into account linguistic knowledge, learning theory, 
the mathematics of codes, and research-based models 
of human motor control, in addition to the "usual" 
range of topics central to interface design for mass- 
market computing. 

• Another basic lesson taught to engineering design 
students is that they generally make poor models of 
their intended customers; in other words they must 
find out what their market wants rather than assuming 
that they can draw useful conclusions from their own 
preferences and needs. This becomes particularly 
clear, for example, to students designing an 
independent eating system for diners with 
quadriplegia. These consumers are so dramatically 
different in their physical capacities that students 
realize that they can’t assume anything about less 
obvious characteristics such as how they trade off 
independence of function against simplicity and 
economy of the device. 

• We also want our students to learn from their 
design-projects courses that the ’’soft" engineering 
considerations — for example comfort, ease of 
learning, and ergonomic efficiency — are in fact 
really hard. Design of products for people with 
reduced capacities makes it particularly clear that these 



considerations must be addressed carefully if an 
invention is to succeed in the marketplace. 

• If a student ends up working for Boeing, for 
example, he/she will be professionally preoccupied 
with "human-machine systems" such as the 
performance of the combined attributes of a human 
pilot and a high speed jet aircraft. He/she will 
discover that what matters is how well the designer 
arranges for a match between the operator’s task and 
the operator’s abilities, a lesson that is taught 
particularly well by designing assistive technology 
since the designer’s attention is drawn to the 
operator's abilities by the fact that they are so 
obviously different from his/her own. 

• Students destined to be involved in product design 
need to learn that a successful product must "win" 
both for its objective performance and for the pleasure 
the user derives from its aesthetics. The importance 
of both of these determinants of success is 
emphasized in design of assistive technology since 
performance is critical since for elimination of 
functional handicaps and aesthetics is also crucial 
since the way a person with a disability perceives 
him/herself and is received by the world is very 
sensitive to the way s/he looks. 

• Finally, we teach design students that one of the 
reasons that the American auto industry had lost 
ground to Japanese manufacturers is that Detroit had 
not paid sufficient attention to the diversity and rapid 
change of tastes, preferences and driving 
circumstances of its intended customers. Assistive 
technology offers students a particularly vivid 
example of the need to design products which are 
modular or field-programmable or customizable on 
the assembly line to accommodate the extreme range 
of needs and abilities that characterizes its market. 

ACKNOWLEDGMENTS 

Funding for preparation and presentation of 
Product Design for People with Disabilities is 
provided by the Biomedical Engineering Department 
of the University of Tennessee Memphis. 

REFERENCES 

1- Rosen, M.J., Editorial, Assistive Technology, 

3(4), 1993. 



Michael J. Rosen 

Department of Biomedical Engineering 
The University of Tennessee, Memphis 
Suite 801 
899 Madison Ave. 

Memphis, TN 38163 
(901)448-7653 
FAX: (901) 448-7387 
mrosen @utmeml .utmem.edu 

247 

June 17-22, 1994 



RESNA ’94 



233 



RR&D’S BAG OF TECHNOLOGY TRANSFER TOOLS 



David L. Jaffe 
VA Medical Center 
Palo Alto, CA 



ABSTRACT 

One of the most challenging goals of the Palo Alto 
VA’s Rehabilitation Research and Development 
Center (RR&D) is the transfer of its laboratory 
prototypes to companies for manufacture. Without a 
purposeful program of technology transfer, devices 
developed here would remain mere research 
curiosities and would not benefit disabled veterans. 

Despite past obstacles to technology transfer, there 
now exist a variety of tools and incentives that can 
promote commercialization of RR&D’s prototype 
devices. 



BACKGROUND 

Commercializing the Center’s rehabilitation products 
is a formidable task since many of them face a limited 
inarket, are costly to develop, are sold to people with 
limited financial means, involve third party payers, 
and are subject to many Federal regulations. For diese 
reasons, investors and manufacturers are often 
hesitant to address this market, even though the need 
for the product may be great. 

For RR&D, the process of commercializing a 
prototype device requires a substantiaJ conunitment of 
manpower and resources. Contacts with potential 
manufacturers must be developed and maintained; 
advanced prototypes and enhancements of the project 
may be required for demonstration; supporting 
documents and promotional materials must be 
developed and produced; and patentability and 
marketability must be investigated. 

OBJECTIVE 

An important part of RR&D’s mission is to develop 
assistive devices which will directly benefit disabled 
veterans. Since the VA cannot by law manufacture 
the devices developed at the Center, it must rely on 
industrial companies to build and market them. The 
goal of technology transfer therefore, is to promote 
RR&D’s prototype devices to commercial businesses, 
move the expertise embodied in the prototype to an 
interested company, and to often work with that 
company to insure a manufacturable and marketable 
product that can serve individuals with disabilities. 

METHOD /APPROACH 

Over the past seven years the repertoire of tools that 
facilitate technology transfer has increased 
dramatically. Despite the existence of these tools. 



there is no simple protocol that is universally 
applicable to every technology transfer situation. In 
each instance, a different complex mix of preparation, 
documentation, demonstration, contacts, and 
serendipity is required. Some of the tools and 
methods available to RR&D are described below. 

VA Technology Transfer History 

A Technology Transfer Section (TTS) has been 
formed within RR&D to accelerate, encourage and 
promote the transfer of technology, including the 
commercialization of RR&D prototype products 
(now numbering thirty-one) through collaboration 
with industry. Tliis group serves to advise, organize, 
supervise, and coordinate all of the Center’s TT 
activities. It also serves as a resource on these 
matters to other investigators within the Center, 
explores and maintains contacts with industry, 
arranges workshops, prepares agreements, and 
carries out negotiations when appropriate. 

TTS has now formulated a strategy for technology 
transfer that comprises three programs: 1) 
Recruitment - identifying potential new products 
from within the Center, 2) Evaluation - identifying 
criteria and methods for screening projects for 
commercial potential, and 3) Availability - 
developing and maintaining manufacturer contacts, 
disseminating information, and negotiating 
agreements or licenses. 

Within TTS a systematic way to evaluate the 
technology transfer potential and commercial 
feasibility of RR&D’s projects has been developed. 
The prototype product’s value in each of the 
following areas is determined: 1) its value to a 
potential licensee including potential market size, 
risks involved, and competitive advantage offered; 2) 
its value to the user or purchaser including 
importance of need, satisfaction of need, and 
alternatives to the product; and 3) its value to the VA 
including concurrence with the Center’s mission and 
existence of champions inside and outside the 
Center. 

A first round of assessments on six projects has 
provided valuable feedback about the evaluation 
process, as well as giving a common reference for 
discussion about specific projects. Project values 
which scored low in the assessment process can now 
be seen as targets for improvement. For example, the 
value to licensee might be rated low because of small 
potential market size; This can provide impetus to 
the investigator to find wider applications for the 
particular technology. 



234 




RESNA ’94 • June 17-22, 1994 



RR&D’s Bag of Technology Transfer Tools 

A Technology Transfer Advisory Board has been 
formed by TTS to provide broad expertise in all 
facets of technology transfer. It consists of twelve 
non-government specialists in the fields of 
technology licensing, patent and Federal law, 
marketing, third party payers, product design, 
venture capital, rehabilitation medicine, 
rehabilitation service delivery, manufacturing, and 
business. The Board members have been helpful in 
suggesting new tactics and continue to provide 
business and marketing advice. 

TTS has developed a working database which 
contains information on all of our projects, as well as 
information on the TT process. In addition, a 
database of VA decision-niakers, funding sources 
and foundations, rehabilitation professionals, 
clinicians, engineers, potential manufacturers of 
RR&D prototypes, students, users, press and media 
people, and interested lay people is being 
maintained. 

To increase the awareness of colleagues, users, 
manufacturers, health professionals, entrepreneurs, 
and others of RR&D’s projects and products, the 
OnCenter newsletter is published two or three times 
a year. It highlights new developments and 
opportunities in technology transfer, reports on 
specific products or projects, and solicits ideas, 
suggestions, and inquiries about RR&D’s work. The 
distribution of OnCenter has lead to numerous 
personal contacts that are the prelude to successful 
technology transfer. Five issues of the newsletter 
have been published and mailed to over 2500 
individuals and organizations in the TTS database. In 
addition, RR&D routinely publishes and 
disseminates a Progress Report every two to three 
years. It describes all projects being undertaken in 
the Center’s four program areas. 

To assist in the technology transfer and subsequent 
commercial availability of promising devices and 
techniques developed at RR&D, the VA has 
established the Technology Transfer Service (VA- 
TTS) in Baltimore, MD. The VA-TTS funds 
commercial prototypes of RR&D’s devices and 
evaluates them clinically within the VA system. 
Devices they approve may be recommended, 
prescribed, and purchased by the VA for use by 
disabled veterans. Three RR&D projects have 
employed this technology transfer procedure. 

Legislation 

The passage of the Technology Transfer Act of 1986 
by the U.S. Congress authorized local Federal 
laboratories to negotiate and enter agreements 
directly with industry for the purpose of 
commercializing government technology. This 
legislation represents a significant change from the 
previous laws which essentially prohibited such 
collaboration. It offered the first real prospect for 
RR&D to participate actively in the 
commercialization of its products. Before that time, 
it was considered a conflict of interest for 



government employees to be involved in the 
commercialization of technologies developed in 
Federal laboratories. 

Funding and Partnering 

The Technology Transfer Act established a 
mechanism called the Cooperative R&D Agreement 
(CRADA) to promote and facilitate the transfer of 
technology from a Federal laboratory to private 
commercialization. It permits a private company to 
fund and work with RR&D toward the development, 
manufacture, and marketing of a specific 
rehabilitation device. 

In addition to CRADAs, RR&D can also enter into 
collaborations involving the exchange of knowledge, 
facilities, or personnel where no money changes 
hands. 

In 1982 Congress established the Small Business 
Innovation Research (SBIR) program to provide 
small business with the opportunity to acquire 
Federal R&D funds to stimulate the development and 
commercialization of technology for public and 
private benefit. Under Phase I of an SBIR 
competitive award, a small business can receive up 
to $100,000 for up to six months to evaluate the 
technical merit and feasibility of an idea, while in 
Phase n they can receive an additional $750,000 for 
up to two years to expand the Phase I effort and 
develop the technology in preparation for 
commercialization. 

Under a funded SBIR project, both a small business 
and RR&D (serving as a consultant) can receive 
funds that support moving a research prototype out 
of the laboratory and into the marketplace. 

The Small Business Technology Transfer program is 
similar to an SBIR, with awards of $100,000 and 
$500,000 for Phase I and Phase II respectively. It is 
intended to stimulate and foster scientific and 
technological innovation through cooperative 
research and development carried out between small 
business concerns and research institutions, foster 
technology transfer between small business concerns 
and research institutions, and increase private sector 
commercialization of innovations derived from 
Federal research and development. The participation 
of a federal laboratory is mandatory in an STTR and 
appears to be an ideal arrangement for a small 
company that desires to commercialize RR&D 
projects. 

The FDA Orphan Drug Medical Device Grant 
Program appears to support technology transfer for 
medical devices for diseases and conditions that 
affect small populations or have little financial 
incentive to research. They will provide grants of up 
to $100,000 a year for up to 3 years for clinical trials 
for existing prototypes and a similar amount over 
two years for projects with additional development. 
Public and private non-profit and for-profit 
organizations are eligible to apply for these awards. 



• June 17-22, 1994 



RESNA ’94 



235 



RR&D’s Bag of Technology Transfer Tools 

Palo Alto Institute for Research and Education 
(PAIRE) 

PAIRE is a non-profit organization within the VA 
Medical Center that can administer funds from 
CRADAs, SBIRs, STTRs, and negotiated contracts 
that would be difficult to handle under existing 
circumstances. It can support the hiring of personnel, 
equipment purchases, and petty cash reimbursements 
required while working under these situations. 

AZTech, Inc. 

AZTech is a not-for-profit, community-based 
enterprise operated by the Rehabilitation Engineering 
Research Center on Technology Evaluation and 
Transfer under a grant from National Institute on 
Disability and Rehabilitation Research. The 
company will evaluate a prototype rehabilitation 
device for commercial potential, identify corporate 
partners and establish business agreements that are 
intended to move the product to market. Inventors 
(including those who use assistive devices), 
companies, and research centers are encouraged to 
participate in this program. A small application fee 
and a working prototype are required. [1] 

Business Gold 

The National Technology Transfer Center (NTTC) 
has developed an electronic bulletin board system 
called Business Gold, as an easy, convenient means 
of accessing information oii the newest federally 
funded technologies. Updated regularly, the online 
system provides a new directory of Federal R&D 
laboratories and new Federal technologies available 
for commercialization, current Small Business 
Innovative Research solicitations, electronic mail 
facilities. Federal R&D information, and related 
announcements. Future enhancements include a 
conference calendar of technology transfer events, a 
publications list, and education and training 
opportunities. The service is free, and there are no 
online or report charges. 

RESULTS 

RR&D now has over eight years experience with 
technology transfer. Some of the lessons learned are: 

1. Tech transfer does not come about quickly, even in 
the best of circumstances. It requires large doses of 
preparation, experience, and chance. 

2. The time from transferring a working laboratory 
prototype to seeing a product on the market can be 
reasonably short if the company is small, adequately 
funded, and smart and thorough in its approach. 

3. Hiring the RR&D developer is often critical to the 
small company’s success. Even though the VA loses a 
valuable investigator, there may be no more effective 
means of fulfilling its mission to transfer its 
technology. 



4. The movement of promising technology to industry 
offers the private sector valuable knowledge, 
expertise, and R&D effort which they might 
otherwise not be able to afford. 

RR&D’s successes include: 

1) arranged for seven Center products to go into 
commercial production; 

2) produced and distributed a technology transfer 
Guidebook authored jointly with a private company; 

3) published five issues of OnCenter aimed at 
technology transfer professionals and potential 
manufacturers; 

4) collaborated with VA Technology Transfer Service 
in Baltimore and with Gallaudet University in 
Washington to put field four Center products into 
field evaluation; 

5) negotiated the following: two Cooperative R&D 
Agreements with private companies; one license 
agreement; a publishing agreement; and two 
agreements with private companies for patents and 
licensing; 

6) filed twelve disclosures of inventions with VA’s 
General Counsel in Washington. 

DISCUSSION 

Bringing a good project idea to the point of 
comniercialization requires a concerted, long-term 
technical effort, several infusions of funding, and 
plenty of patience on the part of RR&D investigators. 

Technology transfer is like gardening: no one can 
build a rose, but a good gardener can produce a better 
one by tilling the soil well, selecting the right seeds, 
getting the bugs out, and doing the right kind of 
fertilization, watering, and pruning. 

REFERENCES 

[1] AZTech Brochure, RERC-TET, University of 
Buffalo, 515 Kimball Tower, Buffalo, NY 14214- 
3079. 

ACKNOWLEDGEMENT 

This woric is supported by the Department of Veteran 
Affairs’ Rehabilitation Research and Development 
Service. 

ADDRESS 

VA Medical Center 
3801 Miranda Ave., Mail Stop 153 
Palo Alto, C'j«i:.94304 
415/493-5000 ext 4480 
jafi’e@roses.stanf ord.edu 



.250 

. Jt. -’V- ■ 



236 



RESNA’94 • June 17-22, 1994 



DESIGN ISSUES FOR PEOPLE WITH SPECIAL NEEDS - 
A STUDY WITH DRIVERS 

Colette Nicolle & Simon J. Richardson 
The HUS AT Research Institute, Loughborough University of Technology 
The Elms, Elms Grove 
Loughborough, Leics LEI 1 IRG, England 



ABSTRACT 

Design issues which are relevant for drivers with 
special needs were identified in a field study as part 
of the TELAID DRIVE II project. These design 
issues were classified under the following headings: 
User and Task Match Aspects, Usability Goals, 

Input Aspects, Output Aspects, and Physical 
Characteristics of tlie Aid. This paper presents an 
overview of developing design guidelines for drivers 
with special needs and describes certain key issues 
which are emerging. The study emphasises tliat 
many of these design rcquii ements apply to all 
drivers, and also to tlie design of many products for 
people willi special needs. Likewise, designing for 
special needs often makes a product easier for 
everyone to use. 



INTRODUCTION 

Drivers with special needs (or DSN) may require 
adaptations to make tlieir driving easier and safer, 
including adaptable control aids (eg hand-controlled 
accelerator or secondary control levers for 
headlights), or mobility and car adaptation aids (eg 
swivel seats and aids to enter or leave tlie car). 
Advanced technology, applications including 
navigation and route guidance, travel and traffic 
information, emergency call, or collision avoidance 
systems, may also enable more people with special 
needs to drive by improving safety and performance, 
or simply through tlie confidence and security it 
gives Uiem. However, a trend in new technology is 
to intioduce alternative modes of input or output, eg 
voice control or sound output. Great care must be 
taken to prevent the future development of cars and 
advanced technology systems taking such a dii*ection 
that those witli perceptual impainnents will become 
drivers with special needs, or tliat the driving task 
will become more difficult tlian before, leading to a 
decrease, not an increase, in driver safety. 

The TELAID project (Telematic Applications for tlie 
Integration of Drivers with Special Needs) is part of 
the Advanced Transport Telematics DRIVE II 
programme of the Commission of the European 
Coimnunities. TELAID's main objective is to apply 
developments in new technology to extend tlie range 
of assistance that can be given to drivers witii special 
needs. 



Using a systems approach, other types of aids were 
also included in the study, as the use of another aid 
might lead to a different type of performance or 
workload. A final product of the project, which 
ended its first phase in December 1993, is the 
development of design guidelines for aids for drivers 
with special needs. This paper presents an overview 
of these guidelines and describes certain key i.^sues 
which are emerging. 

Guidelines exist to assist fitters of car adaptations 
when making technical conversions to a vehicle to 
meet a disabled driver's functional requirements (eg 
5). The 'Draft Code of Practice for In-Vehicle 
Information Systems' (1) covers key issues that must 
be considered when designing advanced technology 
systems that will be used by the driver while driving. 
However, car manufacturers, car adaptation 
manufacturers and research projects often do not 
have the resources to satisfactorily include the wide 
ranging needs of different impairment groups. As 
drivers with special needs are often closer to the 
limit of their abilities, the demands on good design 
become more explicit for them, and pitfalls of 
adaptation solutions may result in even more severe 
consequences (4). Therefore, guidelines are needed 
to ensure that DSN are included in the design 
process. 

METHOD 

Tlie constraints, limitations and requirements of 
DSN were identified in a field study, following the 
Cascade Model' for data collection (3). Two tools 
were developed specifically by the TELAID 
consortium to collect comprehensive data: 

• A comprehensive definition of the driving task 
to serve as a checklist, or structured interview. 
This list of components of tlie driving task 
totalled 1 1 main tasks, 73 sub-tasks and 165 
prompts, to assist the interviewer to cover the 
required level of detail, ranging from steering 
and accelerating to planning a journey. 

• A classification tool to identify the range of 
impairments which should be included in the 
term “Drivers with Special Needs”. The 
classification was based on the effects of the 
impairment, rather than the cause of an 
impairment. Tlie impairment groups included in 
the data collection were visual, reading. 




RESNA ’94 • June 17-22, 1994 



237 



Design Issues for People witJi Special Needs 



hearing, speech, lower limb, upper limb, upper 
and lower limb, upper body, sudden loss of 
control, and cognitive. 

RESULTS 

This study had two major results: 

• It highlighted components of the driving task 
which cause difficulty for each or every 
impairment group. This helped to identify the 
problem areas, driving tasks and target groups 
to investigate in the TELAID simulator tests 
(4). 

• Problems experienced by drivers with special 
needs identified design requirements for 
existing or prospective aids, which formed the 
basis of the TELAID design guidelines for 
aids for drivers with special needs. 

A full description of tliese design requirements can 
be found in TELAID Deliverable No. 6A (2). These 
issues form tlie basis of the TELAID design 
guidelines for aids for DSN, which will soon be 
submitted to die CEC in the final deliverables of the 
project. Only a selection of issues are given below 
because of their relevance not just for drivers, but for 
all people with special needs. They are classified 
under the following headings, cascading from high 
level categories of guidelines which can be broken 
down into lower level guidelines and specifications. 

User and Task Match Aspects 
These design aspects refer to how well the 
functionality of the aid or system meets the specific 
and often changing requirements of the user, and 
also how the user is supported in the performance of 
the required task. Take, for example, "Meeting the 
real needs of the user." With such heterogeneous 
user groups, these needs cover a wide range and may 
also be changing over time, due to degenerative or 
variable illnesses, eiUier over the long tenn or die 
short term. People widi variable illnesses may have 
reduced physical or cognitive abilities on certain 
days, due to anxiety, stiess, or climate. For example, 
an ardiritic driver may have particular difficulty in 
damp weather when using the secondary controls (eg 
for heating or directional signals), and designers 
should consider this issue when chtx)sing the size, 
location, shape and sensitivity of the controls. 

Usability Goals 

These design aspects refer to how well the aid or 
system supports a particular user in achieving 
specified goals in a pai ticular enviroiunent, 
effectively, efficiently, safely and comfortably. Take, 
for example, "Flexibility and Adaptability." To what 
extent can die DSN easily adjust an aid to suit 



individual requirements? Can users easily input or 
interpret output, or adjust the location of the 
input/output device, whether it be in the car or when 
using a control panel in the home? Those with 
mobility impairments may have problems adjusting a 
car seat, especially when full back position is 
required to get in and full forward position is needed 
to drive, and also when other people drive the same 
car. This design issue is also relevant in "smart 
home" technology, when, for example, a product like 
a telephone or remote control needs to be used - and 
accepted - not just by the elderly or disabled person, 
but also by other members of the family. 

Input Aspects 

These design aspects relate to how the user 
communicates with or manipulates the aid or system, 
eg through writing, speaking, or physically operating 
the controls. Take, for example, "Choice of input 
mode and workload on residual abilities." Is the 
driver able to choose the most appropriate mode of 
input to cater to individual needs (eg voice or tactile, 
novice or expert facilities)? The aid or system must 
not impose an overload on residual abilities or on 
one c^ability trying to compensate for another. 
When a paraplegic driver uses hand controls for all 
the primary and secondary driving tasks, do the 
upper limbs have to work excessively to compensate 
for lower limb impairments? The introduction of any 
new technology in the vehicle must consider the 
already heavy workload imposed on that driver. 
Alternative modes of input and output should be 
investigated, for example, voice activation for 
secondary controls, bright sunshine (eg for putting 
down the sun visor), and moisture (for activating the 
windscreen wipers). Users must be able to operate 
controls effectively and safely without using total 
available force or controlling strength. If constantly 
working to the limits of one's capacity, a person with 
special needs would be unable to cope with 
emergency situations, whether it be in the car, in the 
home or anywhere else. 

Output Aspects 

Tliese design aspects relate to how the user receives 
or interprets information from the aid or system, eg 
tlirough reading, listening, receiving tactile 
messages, or viewing graphics on a display. Take, 
for example, "Choice of output mode and workload 
on residual abilities." Is the driver able to choose the 
most appropriate mode of information presentation 
to meet any special requirements (whether it be 
visual, auditory, tactile, or other)? As with the input 
aspects above, the product or system must not 
impose an overload on residual abilities or on one 
capability trying to compensate for another. The 
visual channel, for example, must not become 
overloaded for hearing impaired people, and the 



ERIC 



238 



RESNA^94 



June 17-22, 1994 



Design Issues for People with Special Needs 

tendency to introduce voice output in cars might lead 
to some people becoming drivers with special needs. 

Physical Characteristics of the Aid 
These design aspects are the tangible characteristics 
of the object, eg whether a slippery surface texture 
facilitates or hinders performance of the task. Take, 
for example, "Texture of surfaces." To what extent 
does die texture of the object facilitate or hinder 
performance of the required task? For example, 
wheelchair users find that slippery surfaces make it 
easier to get in and out of the car, and some drivers 
interviewed even carried a plastic carrier bag with 
them to slide in and out. Possible reflections or glare 
on surfaces, controls or displays might also hinder 
performance. Such glare could cause discomfort or 
even further impairment, and must be considered in 
any system development, whether it be in die home 
or in die car. 

DISCUSSION AND CONCLUSIONS 

Problems experienced by DSN have yielded 
recommendations for improvements in design which 
will better utili.se the residual capacities of drivers 
with special needs and optimally support the 
perfonnance of the driving task. These 
recommendations form die basis of design guidelines 
for aids for DSN, a selection of which have been 
discussed above. 

Although diis paper emphasises drivers with special 
needs, diese design requirements should apply to all 
drivers, and likewise to the design of many products 
for people with special needs. However, it is a 
common error to assume that the design 
requirements of elderly or disabled people are 
unique. Many of us may at times, or will in the 
future, experience some difficulty with, for example, 
inserting the key into door locks, fastening seat belts, 
operadng dashboard controls, using the handbrake, 
or parking. Consider the needs, for example, of 
people with young children or heavy .shopping, small 
drivers, or those suffering from even a temporary 
stiff neck. So, if all designers would design for 
everyone, this will help vast numbers of die able- 
bodied population as well. 

ACKNOWLEDGEMENTS 

The authors wish to Uiank the Commission of the 
European Communities for funding the work, and 
the entire TELAID consortium for their 
contributions to this study, especially Manfred 
Dangelmaier and Paulus Vos.sen (Fraunhofer 
Institute/University of Stuttgart), and Bjbm Peters 
(Swedish Road and Traffic Research Laboratories). 



REFERENCES 

1. Institute for Consumer Ergonomics, 
Loughborough, UK. The Design of In-Vehicle 
Information Systems, Code of Practice and 
Design Guidelines (DRAFT). Prepared for the 
UK Department of Transport, April 1993. 

2. Naniopoulos A, Bekiaris E (eds.). Propsective 
Aid Systems for Categories of Needs of DSN, 
EEC/DGVII, DRIVE II project TELAID 
V2032, Deliverable No. 6A, September, 1993. 
(This and other Project deliverables obtainable 
from European Commission Host Organisation, 
CORDIS Customer Service, B.P. 2373, L-1023 
Luxembourg.) 

3. Nicolle C, Ross T and Richardson SJ. 
Identification and Grouping of Requirements for 
Drivers with Special Needs. In Proceedings of 
ECART 2 (European Conference on the 
Advancement of Rehabilitation Technology), 
Stockholm, 26-28 May 1993. 

4. Peters, B and Nilsson, L (1993) Driving 
Performance of DSN (Drivers with Special 
Needs) using hand controls for braking and 
accelerating. 26th International Symposium on 
Automotive Technology and Automation, 
Aachen, Germany, 13-17 September 1993. 

5. Transport and Road Research Laboratory and 
the Institution of Mechanical Engineers. 
Guidelines on the Adaptation of Car Controls 
for Disabled People, prepared for the UK 
Department of Transport. HMSO, 1990. 

Colette Nicolle 

The HUSAT Research Institute 

Loughborough University of Technology 

The Elms, Elms Grove 

Loughborough, Leics LEI 1 IRG, England 

Tel: +44-509-611088 

E-mail: c.a.nicolle@lut.ac.uk 



253 



June 17-22, 1994 



RESNA ’94 



239 



USER INVOLVEMENT IN ASSESSMENT AND USER INFLUENCE IN STANDARDIZATION OF 
CONSUMER PRODUCTS AND ASSISTIVE TECHNOLOGY 



John Gjodeium 

Danish Centre of Technical Aids for Rehabilitation and Education 
Dq>t of Technology, Communication and Special Education 
Aarhus, Denmaifc 



ABSTRACT 

When assistive technology and consumer products are 
st yidardi zed or yAiea they are assessed for general 
suitability, the work is traditionally carried out by 
professionals, e.g. therapists and engineers, vdiereas 
ooDSumers are rarely involved in the process, even 
Plough they are the real experts concerning their own 
lecjiiirements and needs. 

The Danidi Centre of Tedinical Aids for 
R^bidiilitation and Education eqjoys a dose 
OQcqperation with the Dani A organizations of disabled 
perscms, so it was natural to involve users from these 
oigainzatioiis when entering into the fidds of 
assessment and standardization 

Two prefects wm implemented; one on the 
a ss e ssm ent of conventional tdq>hones and one on 

user influence in intonational standardization. In both 

cases vital knowledge was gained from disabled 
users. 

This paper briefly describes methods used in the 
asses sm ent prqject of conventional tel^hones, and 
methods used and experience gained in the project on 
user influence in international 



BACKGROUND 

Traditionally, international standardization work has 
been carried out by maniifactuiers* terfinin^l 
consultants together with tedinical consultants from 
the regulatory and supervisory authorities. This may 
seem natural enough since up until now 
standardization has focussed on the tedinical 
specifications of products in order to aisure product 
uniformity and to safeguard against designs that may 
harm users or damage other equipment. However, 
us^ d^nands as to product hmetionabiUty and easy 
pp^ation continue to rise, and the adenovdedgement 
that "technical” standardization is merely one dement 
in standardization work is gaining ground. At the 
same time, manufacturers are beaming more and 
more aware of the fact that product fimctionability 
and easy operation constitute major competition 
parameters. Conceits such as fimctionability and easy 
operation are of special interest to disabled persons 
because as users, they make greater and more radical 
demmds as to product user->friendliness. 



In close cooperation with the Danish organizations 
of the disabled, the Danish Centre of Tedinical Aids 
for Rehabilitation and Education has participated in 
the work concerning assessment and standardization 
of conventional tdqihones. 



OBJECTIVE 

In cooperation with Danish user organizations, the 
Danish Centre of Technical Aids for Rehabilitation 
and Education will publish a catalogue of all Danish 
standard telephones rated in terms of suitability for 
disabled persons. 

Li addition, in dose association with European user 
organizations, the Danish Centre is involved in the 
standaidizatioii of tel^hone keypads, so 
disabled and dderly persons without difficulty can 
operate a conventional telqihone keypad. 



APPROACH 

In assodatiQn with the Danish organizatiozis of the 
disabled, the Danish Centre has conq)iled an 
^aluati<m profile vdiich has been used to assess the 
individual telephones in terms of suitability for the 
<hsal^ty groiqps mentioned bdow. The evaluation 
proflie consists of iqiprox. 60 assessment items, and 
under each item the following is described; 

*tiie reason for assessing the item in question 
*bow the item is assessed 
-uhich assessing criteria must be observed 
-vMcix disability groups are to assess the item, and 
questions asked 
^comments and observations 
*refer^ces 

A user panel representing the following disability 
groups; the blind, the visually impaired, the speech 
mqwired, the h^ of hearing, the moitally disabled, 
Md the physically disabW, has assessed each 
individu^ telephone and thus provided the basic 
information material for the catalogue. Furthermore, 
the telephones have been examined and rated by an 
interdisciplinary team of technical consultants and 
therapists. 

The individual conqionents and functions have thus 
been assessed upon the basis of concrete knowledge 



'254 

UESNA ’94 • June 17-22, 1994 



240 



User Influeace in Standardization 



of the needs and requirements of the various disability 
groups. Assessment foldings of handsets, keyboar^, 
displays, ringing functions, operation keys, inductive 
coupling, tactile marking etc have subsequ^tly been 
compiled in a joint report which provides a detailed 
description and assessment of each telephone with 
regard to the individual disability groups. 

As a result of this work, the wish for an even 
closer woiking relationship with user organizations 
arose. The opportunity came in connection with a 
project under the European program TIDE 
(Tedmology Initiative for Disabled and Elderly 
People). The project is called HEART (Horizontal 
European Activities in Rehabilitation Technology). In 
a sub-study of this project, the Danish Centre will 
carry out a case study concerning user influence in 
European standardization. In this work, user 
organizations are involved in monitoring and 
controlling the study, so that vsm are able to 
influence the standariization process and thereby 
influence the result. 

In other words the study will result in a useful and 
tq>plicd>le method of ensuring user influ^ce 
throughout the entire process of standardization work, 
at the relevant levels. In addition, the study will result 
in a draft standard of a product which will follow the 
usual procedure, includbg a public ^quiry. 



**persons with reduced movem^t capabilities 
•persons with visual impairments 
•blind persons 

The next step was to hold a training session for the 
r^resentadves to ensure that they had a certain 
knowledge of the problems concerning the 
elaboration of standards. 

After the training session, the actual work could 
begin. A control group was set up, consisting of 
user experts (the trained user rqires^tatives) and 
professional experts in sta n da r d iz ation md disabled 
persons’ accessibility to telecotnmuncations: 
tbenq>ists, terdmical consultants and sociologists. 
The task of the control group was to monitor the 
study to make sure that users have the necessary 
influence on the standardization process and thus 
also on the result 

Ilie control group decided to follow a method 
based on an analysis of already existing products (in 
this case; teIq>hone keypads) and the requirements 
disabled users have in order to operate them with 
the least possible difficulty. 

However, a method based on mock-ups is equally 
applicable and can be used in connection wifo the 
standardization of future products. 



Experience from the work concerning the 
assessment of conventional standard td^hones has 
shown that there is a need for the standardization of 
telq)hone keypads. After a round of discussions with 
relevant organizations of the disabled, it was decided 
that the subject of standardization would be tel^hone 
key size and key spacing, with the purpose of 
preparing a standard so disabled and elderly 
persons without difficulty can use a conventional 
telephone keypad. Also, the subject of standardization 
has been approved by ETSI (European 
Telecommunications Standard Institute). 

The first step of the study consisted in making 
enquiries at relevant Danish disability organizations 
to find out v/bsOitT they could select represmitatives, 
i.e. experts from their organizations who would be 
interested in taking part in the study. (Qualification 
requiremmits included a solid Imowledge of the 
individual disability group’s functional inq)mrments 
and the consequences of these inpairments. 
Furthermore, it was necessary that representatives 
were able to speak English as some of the work 
would take place in an international environment. All 
exp^ises in connection with the study were covered 
by the project. Two representatives from each of the 
following five disability groups were selected; 

- mentally retarded persons 

- persons with uncoordinated movements 



The method will be presented to the European 
organizations of the disabled so that the control 
group may receive relevant cormnents and 
incorporate these in the flnal proposal for ’User 
Influence in Standardization*. 

The aiudysis was carried out in the form of a user 
panel test of 12 different tel^hones. The 12 
telq>hones are rq>resentative of ipprox. 60 different 
tel^hones selected from the European market. 
About 60 persons with specific functional 
impairm^ts took part in the user panel test which 
included analyses of various keypad layout designs. 

The material from the user panel test was then 
processed statistically and provided the background 
materia} for the elaboration of a draft standardization 
concerning telqihone key size and key spacing. 

DISCUSSION 

The project is still in progress and does not end 
until September 1994. The project will be evaluated 
and the methodology applied will be discussed in 
organizations of the disabled, relevant parties of 
standardization work, and standardization bodies. 

Since this study has been conducted as a case study 



C01?¥ AmSMBLl 

ERIC 







RESNA ’94 



Jane 17-22, 1994 



241 



User Influeace in Standardization 



concerning the standardization of a specific product, 
consideration must be given, among other things, as 
to how the method can be used in general in 
standardization work to «isure relevant user 
influence. 

Several conclusions, however, may be drawn already: 

- It is of great importance that users take an active 
part in standardization work. In order to ensure 
this, it is vital that organizations of the disabled are 
aware of standardization tasks so that user 
r^resentatives are motivated and carefully, selected. 

- An introduction to standardization work, e.g. by 
training, is in^>ortant so that user participants may 
gain an overall insight into standardization work and 
procedures. 

- It is necessary that the organizations of the disabled 
regard standardization work as an important task 
and organise themselves so that it is possible for 
them to participate. 

- In future, standardization work must be given 
financial support so that users from organizations of 
the disabled are able to participate when necessary. 

REFERENCES 

Access to telecommunications for people with special 
needs. Recommendations for improving and adapting 
telecommunication terminals and services for people 
with inq>airments. ETSI ETR 029 Human Factors 
(HF). 

Usid>ility Checklist for Telephones. ETSI DTR HF 
3002, version: 8. 

Issues in Telecommunicadon and Disability. Edited 
by Stqrhen von Tetzchner, 1991. CEC, COST 219. 

Use of telecommunication: The needs of people 
with disabilities. European Cooperation in the field of 
Scientific and Technical Researdi. COST 219, 1989. 

Accessible design of consumer products. Guidelines 
for the design of consumer products to increase their 
accessibility to people with disabilities or who are 
aging. Working draft 1.6, December 1991, Trace R 
Sc D C^ter, Madison, USA. 

Uniform Federal Accessibility Standards. 

Aspects of telephone use by people with hearing 
loss. Paul Coverdale BNR. RESNA SIG 8, Special 
Session, Toronto, 1992. 



ACKNOWLEDGEMENTS 

Financial support to the project concerning the 
pr^aration of a catalogue of all Danish standard 
tel^hones rated in terms of suitability for disabled 
persons has been given by the Danish Ministry for 
Communications. 

Financial support to conduct a case study 
concerning user influence in European 
standardization work has been given by the 
European program TIDE (Tedmology Initiative for 
Disabled and Elderly People). 

Our thanks to the Danish Council of Organizations 
of Disabled People, and to the involved 
organizations of the disabled for their kind 
cooperation in both projects. 

Our thanks also to the National Telecom Agency 
and the Danish Standards Association for their 
fruitful cooperation. 

John Gjoderum 

Tedmical Consultant, B.Sc.E.E. 

Damsh Centre of Technical Aids for Rdiabilitation 
and Education 

DepU of Technology, Communication and Special 
Education 

Graham Bells Vej lA 
DK-8200 Aarhus N 
Tel: +45 86 78 37 00 
Fax: +45 86 78 37 30 



25C 









242 



RESNA ’94 



June 17-22, 1994 



SIG-08 
Sensory Aids 




EYEWEAR THAT PRECISELY CONTROLS LIGHT INTENSITY LEVELS AT THE EYE 



David A. Ross, M^.E.E^ M.Ed. 

Gary L. Mancil, O.D^ F.AA.O. 

Atlanta VA Rehabilitation Research and Development Center 
Decatur, Georgia 30033 



ABSTRACT 

The purpose of this project is to research, develop 
and evaluate eyewear that dynamically and 
“instantaneously” restricts the amount of light reach- 
ing the wearer’s eyes to a preset level. Both Liquid 
Crystal (LQ and electrochromic technologies are be- 
ing explored toward this purpose. A first LC engi- 
neering prototype is currently being evaluated, and 
the investigators intend to obtain an electrochrcnnic 
jHototype for evaluation by Fall 1994. Initial proto- 
types are being tested by a small, diverse population 
of subjects. Both clinic^ trials and mobility trials in 
structured outdoor urban settings are being em- 
ployed. Results of these initial trials will be evalu- 
ated relative to c^timizing the device design. Hnal, 
re-designed prototypes wUl be evaluated by a larger 
population in a more rigorous battery of tests. Re- 
sults will be published and provided to potential 
manufacturers. An initial retail price of less than 
$250 is expected. 

BACKGROUND 

Persons with ocular diseases such as retinitis pigmen- 
tosa (RP), albinism, aniridia and achromatopsia have 
extreme problems with varying light conditions and 
can usually function effectively only under the most 
controlled lighting conditions. Other ocular diseases 
such as macular degeneratimi and conditions affect- 
ing the ocular media (e.g., cataracts, comeal dystro- 
]Ay) have varying effects on retinal adulation. Dark 
adulation times as long as 30 minutes are not un- 
usual [1]. 

For all persons (fully sighted as well as persons with 
low vision), a fairly narrow range of overall illumina- 
tion is optimal. Too much or too little light results 
in dramatic reductions of visual acuity and a corre- 
sponding reduction of visual function [2]. However, 
the unimpaired person has a type of visual reserve 
(i.e., more acui^, more field of view, more contrast 
sensitivity, more light/daric adaptation, etc. than the 
minimum required by the visual task) [3] that gives 
this person the ability to maintain functional perfor- 
mance in less than optimal conditions. 

A fully sighted person exiting a darkened theater into 
bright sunlight may squint, shade the eyes, look 
down, etc. until the eyes have adequatdy adapted 
(only a few minutes); however, this person can con- 
tinue to function while ad^ting to the sudden change 
in lighting and will be able to avoid obstacles and 
negotiate curbs and stairs. But the low vision trav- 



eler, who does not possess this res^e, cannot func- 
tion under these circumstances. 

The low vision traveler experiences two primary 
functional vision problems: detection of changes in 
terrain, such as curbs, and adapting to changing light- 
ing conditions [4]. In a recent national survey, low 
vision consumers and their mobUity instmctors rank 
ordered their most serious (mentation and mobility 
problems. “Changing environmental lighting condi- 
tions” was considered the most difficult mobUity sit- 
uation by both consume and instructcn^ [4]. “Drop 
offs,” down curbs and steps, web rqiorted as second 
most difficult. Persons with low vision w^ also 
found to confuse shadows of buildings and mail- 
boxes with curbs and potholes. In toms of func- 
tional mobUity, this confusion resulted in reduced 
travel speed and gait changes based on reactions to 
shadows [S]. 

Constraining the amount of Ught reaching the low vi- 
sion p^^n’s eyes to a narrow, constant range might, 
in and of itself, provide a sufficient increase in acuity 
and contrast sensitivity to distinguish between a 
shadow and a “drop-off.” In a separate article, the 
authors stated that “the use of absorptive sun lenses 
or individuaUy prescribed illumination-ccxitrol de- 
vices may aUeviate some of this difficulty, but would 
not eliminate the variabUity in visual functioning 
brought on by ever-changing [rapidly changing] 
lighting conditions” [4]. 

Light- absorbing lenses are available and prescribed 
in a variety of styles, colors and levels of light 
transmission. But in order to adapt to a variety of 
conditions (bright sun, cloudy, indoor bright, indoor 
dim, fluorescent or incandescent lighting) it is neces- 
sary for low vision consumers to And a variety of ab- 
sorptive lenses and Ulumination controls, and to con- 
stantly change back and forth among them — a cum- 
bersome task at best. Photo-darkening lens coatings 
are also avaUable and do provide a degree of ac- 
C(Hrunodation to changing Ughting conditions. How- 
ever, the photo-chemical prcKesses currently em- 
ployed do not adapt “instantly” to changing lighting 
c(Hiditions, especially when going from bright sun- 
light into shadow — a particularly hazardous situation 
for persons with low vision. Further, the wearer has 
no control over this photo-process and cannot adjust 
it to constrict the amount of light passing through the 
lens to specified level. 

The concept of “ideal retinal illumination level” 
emerges from the physical property of optimum 
ulus threshold for the eye, which is dependent 




244 



RESNA ’94 • June 17-22, 1994 



Light-Controlling Eyewear 



upon a variety of factors (such as ocular health sta- 
tus, age, and task). The dark achq)tation curve — itself 
a statement of retinal stimulus threshold — is affected 
by pupil size, normal aging, stimulus size and loca- 
tion, and the presence of ocular disease. These vari- 
ables modify the eye's ability to respond and adapt to 
changes in retinal illumination [6]. The conse- 
quences of changes in retinal illumination are many, 
including changes in color vision and in contrast sen- 
sitivity function ((TSF) [7]. 

Therefore, the role of an eyewear device that could 
dynamically and ''instantly” adjust to changing light- 
ing conditi(His to maintain an "ideal retinal illumina- 
tion level” would significantly contribute to the me- 
diation of retinal functicHi, especially in the presence 
of ocular disease. 

RESEARCH QUESTIONS 

Two research questions are to be answered by this 
research. 

If: m eyewear device were constructed with the 
following specifications: 

(a) lenses that can be darkened/Ughtened to 
effect controlled light transmission at any light 
level between 2% and 45% transmission with 
the ability to dynamically change per cent 
transmission "instantly” (within SO millisec- 
onds), 

(b) sensors on the eye-side of the lenses moni- 
toring field-averag^ light levels reaching the 
eye, 

(c) a user-adjustable feedback circuit capable 
of constraining the average amount of light 
reaching the weary's eyes to within ± 3% of 
any selected lumens level, 

(e) partially-transmissive brow and side- 
shielding elements, 

(0 polarizing filters on lenses and shielding 
elements oriented to reduce reflected glare, 

(g) an easy fit over existing prescription lenses 
and frames, 

(h) total eyewear (frame & lens) weight of 3 
ounces or less, and a nose-bearing weight of 
no more than 1.5 ounces; 

Then: would members of the low vision population 
be able to wear and adjust (or have a vision 
specialist adjust) this eyewear to a speciHc lu- 
minance level such that the wearing of the de- 
vice afforded the wearer significantly improved 
functional vision as measured by: 

(a) improved acuity, and/or 

(b) increased contrast sensitivity, and/or 

(c) increased field of view, and/or 

(d) better color vision and/or 

(e) improved indoor/outdoor mobility. 



2. If: that the above specified eyewear does indeed 
significantly improve the low vision wearer's 
functional vision; 

Then: what electrical/materials/construction/cos- 
metic design would best serve the overall needs 
of this population? 

METHODOLOGY 

This research is being conducted in six stages: 

1) An initial evaluation of available materials/ 
techniques (LC and electrochromic) will be 
performed. 

2) Two prototypes will be designed and const: acted 
(one LC and one electrochromic) employing the 
most promising of the available materials. To 
meet the weight requirements and minimize the 
cost of these initial prototypes, the electronics and 
batteries will be placed in a small box that will 
clip to the wearer's belt 

3) The two initial prototypes are being tested by 44 
subjects in (a) a battery of clinical vision tests 
(acuity, contrast sensitivity, field of view and 
color perception), (b) functional vision 
(indoor/outdoor reading and recognition tasks, 
and (c) indoor/outdoor functional mobility trials. 
The purpose of these will be to informally evalu- 
ate the validity of the first research question 
stated above. Subject observations and comments 
will also be solicited and recorded. 

4) Initial comments and subject observations and 
comments will be analyzed and sorted by disabil- 
ity categorization and frequency. Final prototype 
design will be based on these results. Choice of 
LC or electrochromic technologies will be made, 
and the specific materials, electronics, sensors, 
frames, etc., will be made based on user re- 
sponses. 

5) The final prototype will be tested by a larger sub- 
ject population (104). These will range in age 
from 55 to 75 years of age and consist of four 
sub-groups: (a) 26 "normal” subjects, (b) 26 sub- 
jects with central vision loss from age-related 
macular degeneration, (c) 26 subjects with cloudy 
ocular media and (d) 26 subjects with rod/cone 
dystrophy. 

6) Clinical and functional infmnation will be com- 
piled into specific recommendations for a manu- 
facturable device. This will be published and 
supplied to potential manufacturers. 

Qualitative elements of the above tests will be de- 
scriptively presented through measures of central 
tendency and dispersion. Tabulation and graphic 
representations will be employed when appropriate. 
The objective data will be ev^uated with multiple 
paired t-tests or repeated measures ANOVA to iden- 
r) effects and interactions of condition by 



RESNA’94 • June 17-22, 1994 



245 



Light-Controlling Eyewear 



device. Subjective data will form the basis for spe- 
cific case studies that will complement the objective 
findings hy identifying specific examples of advan- 
tages and disadvantages of the eyewear system. 

Given the large functional advantage expected 
(greater than 1 standard deviation), the 44 initial 
subjects should provide adequate power (>.99) to 
differentiate between the contributing factors. The 
104 subjects employed for the final study will pro- 
vide more than adequate power (>.99) to differenti- 
ate among factors, and will give adequate power 
(>.85) for differentiating among the four sub-popula- 
tions. 

RESULTS 

As of this writing, most of the development of the 
LC prototype has been completed, and evaluation of 
electrochromic technology is about to commence. 
Eight LC materials and technologies were evaluated 
for optical clarity, dynamic range, spectral distor- 
tions and UV and infrared attenuation. Materials/ 
technologies evaluated included two type of nematic 
molecules, three types of “guest-host” dyes attached 
to nematic molecules with one or two polarizing fil- 
ters, a phase change device and a color switching 
shutter. Of these, the “guest-host” devices were the 
most promising. Working with a manufacturer, re- 
searchers woe able to develop a “guest-host” for- 
mula and polarizo' grating size that yielded a pofect 
45% to 2% transmission range with no measurable 
(^tical or spectral distortitMi. Lenses employing this 
“guest-host” formula have been constructed and in- 
tegrated into frames. 

Light-level sensors have been evaluated and a type 
of solar cell chosen for use as a light sensor. The 
solar cell’s electronic simplicity, linearity and nega- 
tive power requirements made it an ideal choice. 
Hnally, a controlling feedback circuit has been 
developed and tested. The completed device is now 
functional and is being tested in the laboratory for 
q)tical and spectral distortion and proper function- 
ing und» a variety of lighting conditions. At this 
writing the prototype appears to be operating as de- 
signed, except under certain lighting conditions. 
Under these ctxiditions op^tion becomes unstable, 
causing some visible flickering. The cause of this is 
being investigated and investigators expect a solu- 
tion to be implemented in the near future. 

Subject testing is expected to commence in mid 
April, 1994, and some preliminary results should be 
available by June, 1994. 



ultimate result of this work will be an eyewear de- 
sign for persons with low vision that best suits the 
needs of this diverse population. The investigators 
also foresee a mass market appeal for such a deyice, 
in which case the purchase price could drop to less 
than $1(X) (in 1(X),(}()0 unit per year quantities). An 
initial market price between $200 and $250 is pre- 
dicted by manufacturers (for 10,000 unit per year 
quantities). 

REFERENCES 

1. Fraser, K£. (1992) Training the low vision 
patient. Problems in Optometry. Philadelphia: J.B. 
Lif^incott Co. Vol.4, No.l, h>. 72-87. 

2. Luria, S.M. Vision with Chromatic Filters. 
American Journal of Optometry and Archives of 
American Academy of Optometry, 49(10), 1972m 
818-829. 

3. Whittako', S., Lovie-Kitchin, J. 0993). The 
Visual Requirements for Reading. Journal of 
Optometry and Visual Science, 70, pp.54-65. 

4. Smith, AJ.; De I'Aune, W.; Geruscbat, D.R. 
(1992). Low Vision Mobility Problems: 
receptions of O&M Specialists and Persons with 
Low Vision. Journal of Visualimpairment & 
Blindness, 86 (1) 58-62. 

5. Barber, A. 0985). The Effects of Low Vision 
Aids and Traditional versus Non-Traditional 
methods in the Mobility Performance and Stress 
Levels of Low Vision Individuals. Final Report: 
The Orientation and Mobility of Low Vision 
Pedestrians. Philadelphia: Pennsylvania College of 
Optometry. 

6. Davson, Hugh (1976). The Physiology of the Eye. 
Academic Press, New York (3rd Ed), 205-220 

7. Waner, Jdm S., Peterzell, David H., Scheetz, 

A J. Light, Vision and Aging, Optometry & Vision 
Science 67(3), 1990, pp 214-229 

ACKNOWLEDGEMENTS 

Funding for this project was provided by the 
Department of Veterans Affairs, Rehabilitation 
Research and Development Service. 

The Atlanta VA Rehabilitation R& D Center is part 
of the Atlanta VA Medical Center Research Service, 
which provided laboratory space and equipment, 
administrative support and staff resources for this 
research. 



DISCUSSION 

If the initial hypothesis holds true — that is, that re- 
stricting the intensity of light readiing the wearer’s 
eyes to a narrow range will significantly improve 
(optimize) the wearer’s functional vision — then the 

ri 

.4 



i 



David A. Ross 
Senior Biomedical Engineer 
Rehabilitation R&D Center 
Atlanta VA Medical Center 
ife70 Qairmont Road (MS 153) 



RESNA’94 • June 17-22, 1994 



246 



A LIQUID CRYSTAL ADJUSTABLE POWER LENS MAGNIFIER 
FOR PERSONS WITH LOW VISION 



David A. Ross, M.S.E.E^ M.Ed. 

Gary L. Mancil, O.D^ F.A.A.O. 

Atlanta VA RehabOitation Research and Development Center 
Decatur, Georgia 30033 



ABSTRACT 

The purpose of this one-year pilot study is to inves- 
tigate and develop Liquid Crystal (LC) technology 
for use as an Adjustable Power Lens (APL) magni- 
fier fw pCTSons with low vision. This is a very new 
concept, as technology has not previously been 
available for the constniction of an APL. Normally, 
once a lens is formed, its power is fixed because 
power is a function of the lens material and the cur- 
vature of the lens. However, it is now possible to 
control the power of a lens made of specially fomu- 
lated LC materials. Once developed, the power of 
an LC APL could be continuously adjust^ by the 
user to achieve optimal visual ability regardless of 
reading task or changes in visual function 

The specific device outcome of this project is to be a 
single-lens 4-powCT clip-on (jeweler’s loop) magni- 
fier with at least a 6 diopter adjustment range (2.5x 
to 5.5x). More importantly, the result will be a more 
highly developed technology fOT the design and 
implementation of APL devices. Ultimately, such a 
lens may also benefit the much larger peculation of 
persons with presbyopia in the form of a single-lens 
replacement for bi-foe^. 

BACKGROUND 

Statement of the Problem 
The condition of low vision exists when a person's 
best corrected visual acuity (i.e., using traditional 
glasses or contact lenses) is insufficient for the suc- 
cessful performance of daily living activities and/or 
the ability to easily read regular-sized print. In 
America today an estimated 4.3 miUion have this 
problem, and the number is growing rapidly with the 
extended life-expectancies of the aging population 
[1]. Loss of vision functioning is most prevalent 
among elderly persons due to the incidence of visual 
problems relat^ to normal aging and age-related 
visual pathologies such as cataract, macular degen- 
mtion, glaucoma and diabetic retinopathy — the 
most prevalent of visual impairments. Estimates of 
the numbers of elderly persons with significant loss 
of vision function include 13% of non-institutional- 
ized elderly persons. These estimates are signifi- 
cantly higher within older age groups: 16% of those 
75 to 84 years old and 27% of those 85 and older are 
estimated to be visually impaired [2]. 

Low vision results when an eye disease or injury 
occurs resulting in permanent damage to the eyes or 



as the result of some congenital condition which has 
damaged the eyes. In the presence of low vision, 
many of the visual tasks a person must complete in 
the course of their activities of daily living become 
impossible. 

The health care mission of low vision rehabilitation 
is to facilitate functional independence by maximiz- 
ing the usefulness of residual vision. The inability 
to access print is viewed by most as especially debil- 
itating [3], and consequently a common goal of 
visually impaired individuals seeking rehabilitation 
is an improved ability to read. Various multi-disci- 
plinary strategies exist, including prescribing optical 
devices, non-optical devices, electronic devices, and 
teaching emnpensatory skiUs. However, this goal is 
most often addressed through the use of optical 
devices that provide magnification of print to a 
“readable” size. Microscopes (high-powered read- 
ing glasses), magnifiers (hand-held and stand 
models), telescopes arranged for near focus 
(telemicroscopes) and loupes (variations on jewel- 
er’s loupes) are conunonly prescribed. 

But these reading devices have limitations in that 
each is useful for only a single task or a narrow 
range of tasks. The result of this is that the person 
with low vision must maintain a coUection of task- 
specific devices. This is an expensive solution that 
increases the complexity and time spent in the per- 
formance of ordiniary activities and increases the 
amount of training required for the successful 
accomplishment of these tasks. 

Traditional lenses are of fixed power/focus, meaning 
that once a given power is ground into a lens, that 
lens’ power cannot be varied. Lens power is deter- 
mined by manipulating three primary variables: (1) 
the index of refraction (n) of the lens material 
(common materials vary from n= 1.5 to 1.7); (2) the 
base curves ground onto the front and back surface 
of the lens; and (3) (in stronger lenses) the center 
thickness of the lens. While this lens property of 
fixed power is appropriate when used in glas^s for 
correcting common refractive errors (myopia, 
hyperopia and astigmatism), the use of fixed focus 
lenses in low vision optical devices limits their range 
of use. 

For example, a hand magnifier of fixed focus/power 
may work well for print of one size, but for smaller 
print a separate, stronger magnifier may be required. 
The same is true for the other optical devices. Fur- 
thermore, many ocular diseases which result in low 
vision are unst^le so that gradual decline in vision 



261 



RESNA’94 • Jiin« 17-22, 1994 



247 



Liquid Crystal Adjustable Power Lens 



is expected (e.g., macular degeneration) while others 
are characterized by periodic fluctuations in vision 
(e.g., diabetic retinopathy). 

The ability to vary the power/focus of a single opti- 
cal device would be of significant benefit in these 
cases. With the use of such a system printed mate- 
rial could be accessed consistently regardless of 
varying print sizes/quality, day-to-day fluctuations 
in visual function or long-term vision changes. One 
optical device could maintain its usefulness over 
time through various stages of decreasing visual 
function. 

Technical Background 

(Chemists have studied the curiosity of LC matftrials 
since the early 1900’s and have known that when an 
electric field is applied to some of these materials 
their index of reflation changes. However, the idea 
to form a lens of LC material to achieve adjustable 
power was not pursued until 1980, when Dr. Ronald 
Schachar filed a patent for what he called a 
“Multifocal Ophthalmic Lens” [4]. Dr. Schachar 
conceived of a “bi-focal” lens constructed of LC 
materials that would automatically adjust its focal 
length based on the sight distance of viewed objects. 
Working with a cooperating LC manufacturer, he 
was able to devise a first prototype demonstrating 
the feasibility of the concept. This device was con- 
structed on a glass slide in which LC material was 
sandwiched between the slide and +1 diopter glass 
lens (See Figure 1). 



+1 diopter 
glass lens 





•2 to >1 diopter 
LC lens 



glass slide ///y^///A 




Figure 1. Crossectional Dlustration of Schachar 
Prototype LC APL with an effective -1 to 0 
Diopter Range. 



As the index of refraction of the LC material in this 
lens decreases with applied voltage the power of the 
LC lens increases from -2 diopters to -1 diopter, and 
the effective power of the LC APL varies finrn -1 to 
0 diopters. 



One problem was identified with the construction of 
this prototype: optical distortion and discoloration 
were apparent in the thicker areas of the LC material 
(the sides of the prototype). Further investigation 
revealed that this was caused by a difficulty in prop- 
erly aligning the LC molecules when the distance 
between the containing walls was increased. This 
difficulty places a limitation on the maximum power 
(and consequently power range) of an LC lens. 

Another concern relates to optical distortion at 
midrange magnification. Here, maintain ing an even 
electric field throughout the LC material (and thus a 



consistent index of refraction) is complicated by the 
variance in spacing between the slide and the lens. 

Both of the above concerns are being attacked by the 
current research. 

RESEARCH QUESTIONS 

The investigators hypothesize that an LC APL can 
potentially eliminate the need many pmons with 
low vision have to use multiple opti(^ aids. Given a 
member of the low vision population, they hypothe- 
size that an LC APL could be devised with a mid- 
range power and power range that would meet all 
the reading needs of this individual the majority of 
the time — even as this user’s visual function fluctu- 
ated or shifted with time. The following key ques-- 
ticxis to be answered in this pilot study relate to spe- 
cific optical and human factors issues/conditions that 
must evaluated. 

Optical Questions 

1. Can an LC APL be developed with a dioptric 
range of 6 or greater? 

2. Can an LC APL with a range of 6 diopters be 
designed to meet the standards of clarity and li gh t 
transmission required for ophthalmic-quality 
lenses? 

Human Factors Questions 

3. Does a panel of 5 experts in vision rehabilitation 
(having been shown a demonstration of a proto- 
type LC APL) see the potential and practical 
benefit of the LC APL for the low vision popula- 
tion? 

4. Can the LC APL be engineered in a practical and 
useful form in terms of size, weight, mounting 
structure, durability and voltage requirements? 

A “yes” answer to each of these questions will sat- 
isfy the condition that the LC APL has great poten- 
tial use for persons with low vision. 

METHODOLOGY 

This research is being conducted in six stages. First, 
the investigators worked with manufacturers of LC 
material to devise formulations of LC material most 
likely to exhibit large refractive index changes with 
applied voltage. The most likely candidates were 
formulated and enclosed between glass slides for 
evaluation. 

Second, the LC candidates are each being subjected 
to a series of optical tests including evaluations for 
refractive variance with applied voltage, optical dis- 
tortion with applied voltage, evaluations of optical 
clarity and/or optical diffusion and spectral distor- 
tion. 

Hurd, nine LC APL’s will be designed and con- 
structed, selecting LC materials and physical designs 
based on the above results. These will be con- 

62 



RESNA’94 • June 17-22, 1994 



248 




Liquid Crystal Adjustable Power Lens 



stnicted of varying lens thickness and curvature to 
test the limits of power and dioptric range. 

Fourth, the nine LC APLs will be evaluated for 
q>tical clarity, image diffusion, and image distortion 
versus appli^ voltage. 

Fifth, based on the above results, a clip-on LC APL 
will be designed and constructed to e^^ibit the best 
design possible given existing technology. 

Sixth, an expert panel comprised of two low vision 
exp^, two “experienced” low vision consumers 
and a low vision manufacture' will be asked to eval- 
uate the LC lots prototype. The procedure will 
consist of a questions and answes session, a demon- 
stration of tte prototype and a focus group discus- 
sion of the benefits^roblems perceived by the panel. 
Results of the development work and panel conclu- 
sions will then be published. 

RESULTS 

Twelve LC material formulations have been devel- 
(^>ed and sandwiched between 2" by 2" optical glass 
sUdes. The voltage induced refractive variance (An) 
of these samples range from An = 0.23 to 0.29, with 
an ovoall index range from n = 1.S to n = 1.8. 

These have been encapsulated in three thicknesses: 
50 microns, 100 microns and 200 microns. Initial 
observation reveals “cloudiness” (scattering) in 
some samples, but most appear to be optically clear 
and free of distotion. More rigorous testing will 
begin in the near future, and detailed results will be 
available by June, 1994. 

DISCUSSION 

If the initial hypothesis holds true — that is, that an 
LC APL with a dioptric range of 6 or greater can be 
constructed — then the result of this work will be a 
clip-on "loop” magnifier for posons with low vision 
that can be adjusted by the user to easily accommo- 
date varying print sizes and vision changes. The 
cost of this device should be no more than $100, and 
it should replace collections of fixed-power devices. 
In addition, if the results are promising, they could 
lead to the developnent of other devices, such as 
zoom lenses (telescopes), and eventually self-focus- 
ing “bi-focals” for persons with jnesbyqpia 



3. (jenensky, S.M., Berry, S.H., Bikson, T.H. and 
Bikson, T.K. (1979) Visual Environmental 
Adaptation Problems of the Partially Sighted: Final 
Report. Center for the Partially Sighted, Santa 
Monica Hospital Medical Center, CPS-IOO-HEW. 

4 . Schachar, Ronald A. (1981) Multifocal 
Ophthalmic Lens. United States Patent number 
4,300,818. Nov. 17, 1981. 

ACKNOWLEDGEMENTS 

Funding for this project was provided by the 
Dq>artment of Veterans Affairs, RehabUitatimi 
Research and Development Service. 

The Atlanta VA Rehabilitation R& D Cent^ is part 
of the Atlanta VA Medical Center Research Service, 
which provided laboratory space and equipment, 
administrative support and staff resources for this 
research. 

David A. Ross 
Seniw Biomedical Engineer 
Rehabilitation R&D Center 
Atlanta VA Medical Center 
1670 Clairmont Road (MS 153) 

Decatur, GA 30033 




REFERENCES 

1. Nelson KA, Dimitrova E. Severe Visual 
Impairment in the United States and in Each State, 
Journal of Visual Impairment and Blindness. 1993; 
March: 80-85. 

2. Havelik RJ. (1986) Aging in the Eighties: 
Impaired Senses for Sound and Light in Persons Age 
65 and Older. NCHS, Advance Data Vital and 
health Statistics of the National Center for Health 
Statistics, No. 125. 



O 



63 



Jane 17-22, 1994 



RESNA ’94 



249 



AN INSTRUCTION SEQUENCE FOR ELDERLY PERSONS 
WITH MACULAR DEGENERATION USING LOW VISION DEVICES 



Gale R. Watson, MJ£d. 

Atlanta VA Rehabilitation Research and Development Center 
Decatur, Georgia 30033 



ABSTRACT 

Age-related Macular Degeneration is the most 
prevalent cause of untreatable visual impairment in 
this country. Elderly persons with macular 
degeneration may be prescribed low vision devices 
(hand-held or stand magnifiers, spectacle mounted 
magnifiers or telescopes, monoculars or video 
nnagnifiers) in order to perform activities of daily 
living. However, unless the individual is able to 
develop and maintain the underlying visual skills 
necessary, the low vision devices may be rejected. 
This presentation will describe the functional 
implications of age-related macular degeneration, 
the development of visual skills, and how the visual 
skills relate to the successful use of low vision 
devices. 



BACKGROUND 

Extrapolating from the Statistical Brief # 36 "Severe 
Visual Impairment in the United States and in Each 
State, 1990" (Nelson and Dimitrova, 1993) there 
were an estimated 6.07 persons per 1000 persons 
who are visually impaired in the 0-54 age group, 
while an astonishing 104.1 persons in 1000 persons 
are visually impaired persons in the 55 + age group. 
The most dramatic increase is in the "old-old 
population," there one finds an estimated 210.6 
persons per 1000 persons with visual impairment for 
those 85 years old and older. The most prevalent 
causes of visual impairment in this country are age- 
related: macular degeneration, diabetic retinopathy, 
glaucoma and cataract. Age-related macular 
degeneration is manifested by changes is the retinal 
cells of the macula. The macula is the small area of 
the retina which is responsible for fine detail vision. 
At present only two known treatments hold promise 
for the prevention of age-related macular 
degeneration. Dietary zinc has been associated with 
a slowing of the loss of acuity after macular 
degeneration has been manifested in a group of 



persons with ARMD compared to a control group 
who took a placebo (Newsome, Swartz, Leone, 
Elson, Miller, 1988). Protecting the eyes from blue, 
violet and UV radiation by the use of absorptive 
lenses has also been suggested to prevent the 
development of macular degeneration (Goodlaw, 
1991). A small proportion of non-exudative macular 
degeneration advances to the disciform state which 
involves subretinal neovascularization and the 
collection of fluid or blood under the retinal pigment 
epithelium (Dressier, Bressler and Fine, 1988). 

Laser photocoagulation can stabilize vision loss, 
however, damaged vision is not restored. Macular 
degeneration results in permanent loss of central 
vision. Peripheral vision remains intact if there is no 
other pathology. Thirty percent of Americans over 
the age of sixty five have age-related macular 
degeneration (Miller, 1979). It is the leading cause 
of new visual impairment, and its incidence 
increases with age. There is a greater prevalence 
among women (Ferris, 1982). 

STATEMENT OF THE PROBLEM: The end 
result of age-related macular degeneration is usually 
a dense scotoma. Symptoms of macular 
degeneration which has resulted in visual 
impairment include difficulty with reading, inability 
to recognize faces, and the distortion or 
disappearance of the visual field straight ahead. 
Individuals with this pathology will usually have 
developed a strongly preferred eccentric viewing 
position (Cummings, Whittaker, Watson and Budd, 

1985) , though they may not always be aware that 
there is a scotoma present. 

APPROACH: Instruction in the use of vision and 
visual skills has been stressed by many authors 
(Backman and Inde, 1976, Jose, 1983, Faye, 1984, 
Collins, 1987, Stelmack et al, 1987), but research 
which substantiates the efficacy of the rehabilitation 
of vision is sparse. Instruction in the use of devices 
was considered vital to successful device use in a 
follow-up study with individuals who had visual loss 
due to macular degeneration (Nilsson and Nilsson, 

1986) . Goodrich, Mehr, Quillman, Shaw and Wiley 
(1977) found that duration and speed of reading 
increased for users of low vision devices during 
instruction. Instruction was found to be important in 
the use of video magnifiers (Goodrich, et al, 1980). 

IMPLICATIONS: Two studies of instruction for 
persons with macular degeneration in the use of 
vision are notable, however, because of the use of 
control groups which did not receive rehabilitation 




June 17-22, 1994 



250 



RESNA ’94 



Instruction Sequence for Elderly 



instruction. With this research methodology, the 
efHcacy of rehabilitation instruction becomes more 
pronounced. Nilsson (1990) provided instruction in 
visual skills, stressing eccentic viewing, and use of 
low vision devices to a group of individuals with 
macular degeneration. A control group did not 
receive rehabilitation instruction, but the optical 
characteristics of their low vision devices were 
demonstrated. After one month, the group receiving 
instruction achieved significantly better results in 
reading, watching television, and writing letters than 
the control group. The mean reading speed for the 
group receiving instruction was 75.5 words per 
minute, as opposed to 22.6 words per minute for the 
control group. Watson, Wright and De I'Aune (1992) 
studied instruction in visual and cognitive strategies 
for reading with low vision devices for individuals 
with macular degeneration. Their sample was 
divided into three groups, an instruction group 
receiving lessons from a low vision reading 
specialist, a practice group reading structured 
practice lessons at home, and a control group who 
received no intervention. While the instruction 
group scored highest on a reading evaluation, the 
practice group also improved. Both groups achieved 
significantly better results than the control group 
which received no intervention. 

DISCUSSION: Elderly persons with macular 
degeneration will require specialized instruction in 
the use of their visual skills in order to maximize 
success in using vision and low vision devices for 
tasks that increase independence and quality of life. 
This presentation will delineate the functional 
implications of age-related macular degeneration on 
activities of daily living, background research on the 
efHcacy of instruction in visual skills and the use of 
low vision devices, and provide the audience with 
basic guidelines for instruction in the use of visual 
skills which must be developed by persons with 
macular degeneration to maximize residual vision, 
with low vision devices. 

REFERENCES 

Nelson KA, Dimitrova E. Severe Visual Impairment 
in the United States and in Each State, Journal of 
Visual Impairment and Blindness. 1993; March: 
80-85. 



Cummings, R., Whittaker, S., Watson, G., Budd, J. 
(1985) Scanning Characters and Reading With a 
Central Scotoma. American Journal of Optometry 
and Physiological Optics. 62, 833-843. 

Faye, E. E. (1984). Clinical Low Vision, Little, 
Brown and Company, Boston. 

Goodlaw,E. (1991) Preventing Cataracts and Age- 
Related Maculaopathy. Journal of Vision 
Rehabilitation, 5,2. p. 1-8. 

Goodrich, G., Mehr, E., Quillman, R., Shaw, H., 
Wiley, J. (1977) Training and Practice Effects in 
Performance with Low Vision Aids: A Preliminary 
Study. Am. J. Of Opt & Psysio. Opt. 54,5. 

Goodrich, G„ Mehr, E., Darling, N. (1980) 
Parameters in the Use of CCTVS and Optical Aids. 
Am. J. of Opt & Physio. Opt 57, 881. 

Greig, D. E„ West, M. L., and Overbury, O. (1986). 
Successful use of low vision aids: Visual and 
psychological factors. Journal of Visual Impairment 
and Blindness^ December, 985-988. 

Newsome, D.A., Swartz, M., Leone, N.C., Elston, 
R.C., Miller, E. (1988) Oral Zinc in Macular 
Degeneration . Arch Ophthalmol .106: 1 92- 197. 

Stelmack, J., Stelmack, T., Fraim, M., Warrington, J. 
(1987) Clinical Uses of the Pepper Visual Skills for 
Reading Test in Low Vision Rehabilitation. Am. J. 
Optom. Physiol. Opt 64,829-831. 

Nilsson, U. (1990). Visual Rehabiltation with and 
Without Educational Training in the Use of Optical 
Aids and Residual Vision. A Prospective Study of 
Patients with Advanced Age-Related Macular 
Degeneration. Clinical Vision Science. 6 p. 3-10. 

Watson, G., Wright, V., De TAune, W. (1992) The 
Efficacy of Comprehension Training and Reading 
Practice for Print Readers with Macular 
Degeneration. Journal of Visual Impairment and 
Blindness. (86) 1,37-43. 

ACKNOWLEDGEMENTS 



Backman, O., Inde, K. (1976) Low Vision Training. 
Malmo, Sweden: Liberhemods. 

Bressler, N.M. Bressler, S.B., Fine, S.L. Age-related 
Macular Degeneration, Survey of Ophthalmology 
1988, 32-375-413.. 

Collins (1987) Coping with the Rising Incidence of 
Partial Sight, Optometry Today 27, 772-779. 



Funding for this presentation provided by the 
Department of Veterans Affairs, Rehabilitation 
Research and Development Service. 

Gale R- Watson, M.Ed. 

Research Health Scientist 
Rehabilitation R&D Center 
Atlanta VA Medical Center 
1670 Clairmont Road (MS 153) 

Decatur, GA 30033 



0 c'i265 

k ^ v..f =■* 

RESNA ’94 • June 17-22, 1994 



251 



EVALUATION OF A ROBOTIC FINGERSPELLING HAND 



Judith E. Harkins, Elizabeth Korres, and Norman S. Williams, Gallaudet University, Washington, DC 
William Haiwn, A. I. du Pont Institute, Wilmington, DE 
David L. Jaffe, Palo Alto Institute for Research and Education, Palo Alto, CA 



ABSTRACT 

A robotic fingerspelling hand has been proposed as 
an alternative to braille computer peripherals for 
deaf-blind people whose native language is 
American Sign Language and whose onset of 
blindness is after childhood. An evaluation of such 
a prototype with eight deaf-blind evaluators was 
conducted. The device conformed to many 
desirable specifications for such a device, and could 
fihgerspell reasonably clearly for sighted people. 
However, the intelligibility of the device was not 
high enough for practical use in tactile reception. 
The difficulty of producing clear fingerspelling 
through a robotic hand may have been 
underestimated in the past, and a very high level of 
intelligibility would be required in order to 
rationalize this technology for deaf-blind people. 

BACKGROUND 

A robotic fingerspelling hand has been proposed as 
an alternative to braille computer peripherals for 
deaf-blind people whose native language is 
American Sign Language and whose onset of 
blindness is after childhood. Several development 
efforts have resulted in prototypes (Laenger & 
Peel, 1978; Beeson, 1981; Gilden & Jaffe, 1986). 
However, past prototypes could not move fluidly 
from letter to letter in the manner of normal 
fingerspelling, and this limitation precluded 
intelligibility evaluations with the intended users of 
the device. 

In 1992-93, two fingerspelling-hand prototypes 
based on the design of Dexter II were produced for 
evaluation. Software was developed to maximize 
the clarity of the devices’ fingerspelling and to 
incorporate the inter-letter transitions. This paper 
reports on the user evaluation that concluded this 
project. The reader is referred to Jaffe et al. 
(19^) for a description of the hardware design. 

METHODS 

Evaluators were recruited through local associations 
and electronic mail networks. 



Baseline speeds for receptive fingerspelling were 
first measured by having a skilled signer fingerspell 
in time with a computerized metronome. 
Evaluators signed and/or spoke their responses to 
each item. Nonsense strings and short sentences 
were used as stimuli. 

When the prototype was ready for evaluation, two 
sessions of 90 minutes to two hours were conducted 
with individual evaluators. Sessions combined 
guided practice, intelligibility tests, and interviews. 

At the time of the evaluation, the prototypes had 
e^rienced several failures each. Because of the 
fragility of the devices, evaluators were not invited 
to subject it to field conditions. Evaluators did not 
independently operate the prototype. 

During both sessions, evaluators’ comments while 
using the hand were recorded. Interviews were 
conducted in American Sign Language at the end 
of both sessions. 

RESULTS 

The sample was composed of eight deaf-blind 
evaluators. All evaluators of the prototype were 
theoretically candidates for such a device. Half of 
the evaluators were proficient in braille, and half 
were not. Evaluators ranged in age from 19 to 60, 
with a median age of 38. Seven of the evaluators 
became deaf before one year of age, and one had 
onset of hearing loss at age 12. Onset of vision loss 
was congenital for two evaluators; the remaining six 
began to experience vision loss at age 10 or later. 
All reported themselves to be fluent signers, but 
three of the eight said they were only fair in 
receptive fingerspelling. (These three had 
experienced vision loss relatively recently.) 

Evaluators’ receptive fingerspelling abilities varied 
widely. Estimated receptive speed, as estimated by 
the baseline measures using human fingerspellers, 
ranged from one to seven characters per second for 
sentence material, and .7 to six characters per 
second for nonsense strings. 



RESNA’94 • June 17-22, 1994 



252 



Evaluation of a Robotic Fingerspelling Hand 

Table 1 summarizes evaluators’ responses to the 
robotic hand prototype. The data are based on 
alphabets displayed in random order in both 
sessions. Evaluators were given as long a time as 
needed to identify each letter. (Evaluators 
sometimes offered more than one response, as they 
were uncertain. Table 1 reflects the evaluators’ 
initial response to each letter.) 



Table 1 

Intelligibility of Individual Letters 



Accuraev Level 


Letters 


90% or higher 


B, C, F, I, L, 
M, N, R, T, W 


80%- 89% 


D, K, O, Z 


70% - 79% 


J, Q.V 


60% - 69% 


A, G, S, U 


Below 60% 


E, H, P, X, Y 


Note: N e 30 trials for each letter 



Eight deaf4>lind evaluators (Seven @ four iterations 
and one @ two iterations) 

Overall, accuracy of reception of individual letters 
did not improve with practice over the two sessions. 
About half (49%) of 80 simple sentences (ten 
sentences read by eight evaluators) were received 
without error. Only two evaluators accurately read 
more than half of ten simple sentences produced by 
the robotic hand. 

The most successful reader was one who changed 
her normal hand position for receptive 
flngerspelling, and addressed the hand from the 
front. The other evaluators approached the hand 
by placing their palms on the back of the robotic 
hand, in the manner used for human 
sign/fingerspelling reception by many deaf-blind 
people (not all). These eviduators were not 
comfortable with changing positions. Deaf-blind 
people frequently make use of information from the 
movement of the bones and knuckles in the back of 
the hand, and these features were absent in the 
robotic prototype. Evaluators sometimes used both 
hands to examine the device. 



Users found the size and lack of vibration to be 
acceptable. They were able to use the speed 
control knob. There were many suggestions for 
changes in the materials used for the prototype, to 
make it feel more like a human hand. None of the 
evaluators felt that the device was intelligible 
enough for practical use. 

DISCUSSION 

The prototypes produced for this project were very 
successful from several perspectives. The prototypes 
fingerspelled fluidly from letter to letter. Positive 
qualities^ of the devices include their small size, 
portability, quiet operation, negligible vibration, use 
of stand^d 110-volt AC power source, use of 
readily available off-the-shelf components, ease of 
operation, speed control knob, software designed 
for easily editing tables, and consistent output of 
fingerspelling. The prototype won an award from 
a design magazine in 1993. 

The robotic flngerspelling hand performed 
reasonably well for visual reception of fingerspelled 
letters. Unfortunately, the results of the evaluation 
indicate that a device that appears visually to 
flngerspell acceptably well will not necessarily 
remain intelligible during tactile reception. For 
example, the weight of a person’s hand on the 
device can have an effect on hand performance. 

Certain improvements in the flngerspelling could be 
made by software changes. These might include 
work on some of the collision problems (e.g., 
transitions to Y). However, thumb collisions are 
difficult to avoid because the device does not have 
a feedback mechanism to give precise control of the 
thumb. 

Other suggested improvements would necessitate 
hardware changes: 

• The "down” position of the hand was intended 
to simulate the wrist flexion used in the letters 
G, H, P, and Q. The device’s movement was 
imperceptible to many users. The movement 
should be exaggerated either in length or angle, 
and perhaps in both; and timing with concurrent 
hand movements needs to be improved. 

• The basic motion of each finger is a continuous 
contraction. The fingers curl from an upright 
position into a flst. This makes it impossible to 
form an accurate K, P, or E. 




RESNA ^94 • June 17-22, 1994 



253 



Evaluation of a Robotic Fingerspeiling Hand 



^ The index finger and middle finger do not touch 
when both are extended, and thus a clear U 
cannot be made. 

In combination, these problems resulted in too 
many non-standard features in the fingerspelling 
output, so that the evaluators did not find the 
output acceptable for real usage. 

The prototypes initially experienced failures as a 
result of breakage of carbon fiber. Some of the 
problem was alleviated by better packaging for 
shipping, but carbon fiber should be replaced with 
some other material if developers continue to work 
on this type of device. One of the prototypes, after 
a period of repeated repairs, performed reliably 
through at least 25 hours of user testing, and is still 
working well. The other copy was held in reserve 
in case of failure by the first device, and was not 
tested in the same way. 



profitably spent on learning braille. Braille’s other 
advantages include the possibility of displaying lines 
of text (rather than one letter at a time) and the 
ability to display punctuation, numbers, and word 
boundaries unambiguously. 

REFERENCES 

Beeson, W. E. (1981). Design and development of 
a microcomputer controlled mechanical hand for 
deaf-blind communication. Unpublished master’s 
thesis in mechanical engineering. University of 
Oklahoma Graduate College. 

Gilden, D. & Jafife, D. L. (1986). Dexter-a 
helping hand for communicating with the deaf- 
blind. RESNA *86, Employing Technology, 
Proceeding;: of the RESNA Ninth Annual Conference 
on Rehabilitation Technology (pp. 49-51). 
Washington, DC: RESNA. 



CONCLUSIONS 

The robotic fingerspelling prototypes produced by 
PAIRE and ASEL conformed to many of the 
functional and technical specifications developed 
under an earlier task of this project. However, the 
intelligibility of fingerspelling was too low for the 
device to be useful to intended users— those deaf- 
blind people who are more fluent in sign than in 
braille. 

The rationale for the fingerspelling hand is that 
people who are deaf throughout life and who 
become blind as a result of Usher Syndrome or 
other causes might learn to receive text through a 
robotic hand more readily than they woidd learn 
braille. This may or may not be true, but cannot 
be said to have been supported by the results of 
this project. The deaf-blind people who were the 
most accurate readers of the robotic prototype were 
those who were already proficient in braille. To 
those who were in their earlier stages of deaf- 
blindness, and learning tactile fingerspelling, the 
distorted spelling of the hand was a source of 
frustration and discomfort. 

Based on these results, the staff conducting the 
evaluation conclude that a robotic fingerspelling 
hand can be rationalized only if the receiver can 
understand it as accurately and easily as human 
fingerspelling is understood. Otherwise, it would 
seem that the deaf person’s efforts to learn a new 
way of receiving information could be more 



RESNA ’94 



Jaffe, D. L., Harwin, W. S., and Harkins, J. E. 
(1993). The development of a third generation 
fingerspelling hand. Proceeding of the RESNA 
16th Annual Conference^ pp. 161-163. 

Laenger, C. J. & Peel, H. H. (1978). Further 
development and test of an artificial hand for 
communication with deaf-blind people. San 
Antonio, TX: Southwest Research Institute. 

39th Annual Design Review, Best of Category - 
Concepts (1993). I,D, Magazine: The International 
Design Magazine^ June/July, 1993, pp. 166-167. 

ACKNOWLEDGMENTS 

Funding for this project was provided by the 
National Institute on Disability and Rehabilitation 
Research (grant number H113G80189-90) and 
Gallaudet University. 

Judith E. Harkins, Ph.D. 

Technology Assessment Program 
Gallaudet Research Institute 
800 Florida Avenue, NE 
Washington, DC 20002 
jeharkins@gallua.gallaudet.edu 
(202) 651-5257 (voice/TTY) 

(202) 651-5476 (fax) 




June 17-22, 1994 



254 



SURVEY OF DEAF-BLIND TECHNOLOGY NEEDS 



Deborah Gilden, Ph.D. 

The Smith-Kettlewell Eye Research Institute, San Francisco, CA 



ABSTRACT 

A Deaf-Blind Technology Needs Questionnaire was 
designed and administered to deaf-blind individuals 
throughout the United States. The results of 
responses from 62 individuals indicate that while 
some specific new devices are needed, the majority 
of devices requested already exist, but the consumers 
are either unaware of them or are unable to obtain 
them. Some o